1 /* Interface between the opcode library and its callers.
2 Written by Cygnus Support, 1993.
3
4 The opcode library (libopcodes.a) provides instruction decoders for
5 a large variety of instruction sets, callable with an identical
6 interface, for making instruction-processing programs more independent
7 of the instruction set being processed. */
8
9 #ifndef DISAS_DIS_ASM_H
10 #define DISAS_DIS_ASM_H
11
12 #include "qemu/bswap.h"
13
14 typedef void *PTR;
15 typedef uint64_t bfd_vma;
16 typedef int64_t bfd_signed_vma;
17 typedef uint8_t bfd_byte;
18 #define sprintf_vma(s,x) sprintf (s, "%0" PRIx64, x)
19 #define snprintf_vma(s,ss,x) snprintf (s, ss, "%0" PRIx64, x)
20
21 #define BFD64
22
23 enum bfd_flavour {
24 bfd_target_unknown_flavour,
25 bfd_target_aout_flavour,
26 bfd_target_coff_flavour,
27 bfd_target_ecoff_flavour,
28 bfd_target_elf_flavour,
29 bfd_target_ieee_flavour,
30 bfd_target_nlm_flavour,
31 bfd_target_oasys_flavour,
32 bfd_target_tekhex_flavour,
33 bfd_target_srec_flavour,
34 bfd_target_ihex_flavour,
35 bfd_target_som_flavour,
36 bfd_target_os9k_flavour,
37 bfd_target_versados_flavour,
38 bfd_target_msdos_flavour,
39 bfd_target_evax_flavour
40 };
41
42 enum bfd_endian { BFD_ENDIAN_BIG, BFD_ENDIAN_LITTLE, BFD_ENDIAN_UNKNOWN };
43
44 enum bfd_architecture
45 {
46 bfd_arch_unknown, /* File arch not known */
47 bfd_arch_obscure, /* Arch known, not one of these */
48 bfd_arch_m68k, /* Motorola 68xxx */
49 #define bfd_mach_m68000 1
50 #define bfd_mach_m68008 2
51 #define bfd_mach_m68010 3
52 #define bfd_mach_m68020 4
53 #define bfd_mach_m68030 5
54 #define bfd_mach_m68040 6
55 #define bfd_mach_m68060 7
56 #define bfd_mach_cpu32 8
57 #define bfd_mach_mcf5200 9
58 #define bfd_mach_mcf5206e 10
59 #define bfd_mach_mcf5307 11
60 #define bfd_mach_mcf5407 12
61 #define bfd_mach_mcf528x 13
62 #define bfd_mach_mcfv4e 14
63 #define bfd_mach_mcf521x 15
64 #define bfd_mach_mcf5249 16
65 #define bfd_mach_mcf547x 17
66 #define bfd_mach_mcf548x 18
67 bfd_arch_vax, /* DEC Vax */
68 bfd_arch_i960, /* Intel 960 */
69 /* The order of the following is important.
70 lower number indicates a machine type that
71 only accepts a subset of the instructions
72 available to machines with higher numbers.
73 The exception is the "ca", which is
74 incompatible with all other machines except
75 "core". */
76
77 #define bfd_mach_i960_core 1
78 #define bfd_mach_i960_ka_sa 2
79 #define bfd_mach_i960_kb_sb 3
80 #define bfd_mach_i960_mc 4
81 #define bfd_mach_i960_xa 5
82 #define bfd_mach_i960_ca 6
83 #define bfd_mach_i960_jx 7
84 #define bfd_mach_i960_hx 8
85
86 bfd_arch_a29k, /* AMD 29000 */
87 bfd_arch_sparc, /* SPARC */
88 #define bfd_mach_sparc 1
89 /* The difference between v8plus and v9 is that v9 is a true 64 bit env. */
90 #define bfd_mach_sparc_sparclet 2
91 #define bfd_mach_sparc_sparclite 3
92 #define bfd_mach_sparc_v8plus 4
93 #define bfd_mach_sparc_v8plusa 5 /* with ultrasparc add'ns. */
94 #define bfd_mach_sparc_sparclite_le 6
95 #define bfd_mach_sparc_v9 7
96 #define bfd_mach_sparc_v9a 8 /* with ultrasparc add'ns. */
97 #define bfd_mach_sparc_v8plusb 9 /* with cheetah add'ns. */
98 #define bfd_mach_sparc_v9b 10 /* with cheetah add'ns. */
99 /* Nonzero if MACH has the v9 instruction set. */
100 #define bfd_mach_sparc_v9_p(mach) \
101 ((mach) >= bfd_mach_sparc_v8plus && (mach) <= bfd_mach_sparc_v9b \
102 && (mach) != bfd_mach_sparc_sparclite_le)
103 bfd_arch_mips, /* MIPS Rxxxx */
104 #define bfd_mach_mips3000 3000
105 #define bfd_mach_mips3900 3900
106 #define bfd_mach_mips4000 4000
107 #define bfd_mach_mips4010 4010
108 #define bfd_mach_mips4100 4100
109 #define bfd_mach_mips4300 4300
110 #define bfd_mach_mips4400 4400
111 #define bfd_mach_mips4600 4600
112 #define bfd_mach_mips4650 4650
113 #define bfd_mach_mips5000 5000
114 #define bfd_mach_mips6000 6000
115 #define bfd_mach_mips8000 8000
116 #define bfd_mach_mips10000 10000
117 #define bfd_mach_mips16 16
118 bfd_arch_i386, /* Intel 386 */
119 #define bfd_mach_i386_i386 0
120 #define bfd_mach_i386_i8086 1
121 #define bfd_mach_i386_i386_intel_syntax 2
122 #define bfd_mach_x86_64 3
123 #define bfd_mach_x86_64_intel_syntax 4
124 bfd_arch_we32k, /* AT&T WE32xxx */
125 bfd_arch_tahoe, /* CCI/Harris Tahoe */
126 bfd_arch_i860, /* Intel 860 */
127 bfd_arch_romp, /* IBM ROMP PC/RT */
128 bfd_arch_alliant, /* Alliant */
129 bfd_arch_convex, /* Convex */
130 bfd_arch_m88k, /* Motorola 88xxx */
131 bfd_arch_pyramid, /* Pyramid Technology */
132 bfd_arch_h8300, /* Hitachi H8/300 */
133 #define bfd_mach_h8300 1
134 #define bfd_mach_h8300h 2
135 #define bfd_mach_h8300s 3
136 bfd_arch_powerpc, /* PowerPC */
137 #define bfd_mach_ppc 0
138 #define bfd_mach_ppc64 1
139 #define bfd_mach_ppc_403 403
140 #define bfd_mach_ppc_403gc 4030
141 #define bfd_mach_ppc_e500 500
142 #define bfd_mach_ppc_505 505
143 #define bfd_mach_ppc_601 601
144 #define bfd_mach_ppc_602 602
145 #define bfd_mach_ppc_603 603
146 #define bfd_mach_ppc_ec603e 6031
147 #define bfd_mach_ppc_604 604
148 #define bfd_mach_ppc_620 620
149 #define bfd_mach_ppc_630 630
150 #define bfd_mach_ppc_750 750
151 #define bfd_mach_ppc_860 860
152 #define bfd_mach_ppc_a35 35
153 #define bfd_mach_ppc_rs64ii 642
154 #define bfd_mach_ppc_rs64iii 643
155 #define bfd_mach_ppc_7400 7400
156 bfd_arch_rs6000, /* IBM RS/6000 */
157 bfd_arch_hppa, /* HP PA RISC */
158 #define bfd_mach_hppa10 10
159 #define bfd_mach_hppa11 11
160 #define bfd_mach_hppa20 20
161 #define bfd_mach_hppa20w 25
162 bfd_arch_d10v, /* Mitsubishi D10V */
163 bfd_arch_z8k, /* Zilog Z8000 */
164 #define bfd_mach_z8001 1
165 #define bfd_mach_z8002 2
166 bfd_arch_h8500, /* Hitachi H8/500 */
167 bfd_arch_sh, /* Hitachi SH */
168 #define bfd_mach_sh 1
169 #define bfd_mach_sh2 0x20
170 #define bfd_mach_sh_dsp 0x2d
171 #define bfd_mach_sh2a 0x2a
172 #define bfd_mach_sh2a_nofpu 0x2b
173 #define bfd_mach_sh2e 0x2e
174 #define bfd_mach_sh3 0x30
175 #define bfd_mach_sh3_nommu 0x31
176 #define bfd_mach_sh3_dsp 0x3d
177 #define bfd_mach_sh3e 0x3e
178 #define bfd_mach_sh4 0x40
179 #define bfd_mach_sh4_nofpu 0x41
180 #define bfd_mach_sh4_nommu_nofpu 0x42
181 #define bfd_mach_sh4a 0x4a
182 #define bfd_mach_sh4a_nofpu 0x4b
183 #define bfd_mach_sh4al_dsp 0x4d
184 #define bfd_mach_sh5 0x50
185 bfd_arch_alpha, /* Dec Alpha */
186 #define bfd_mach_alpha 1
187 #define bfd_mach_alpha_ev4 0x10
188 #define bfd_mach_alpha_ev5 0x20
189 #define bfd_mach_alpha_ev6 0x30
190 bfd_arch_arm, /* Advanced Risc Machines ARM */
191 #define bfd_mach_arm_unknown 0
192 #define bfd_mach_arm_2 1
193 #define bfd_mach_arm_2a 2
194 #define bfd_mach_arm_3 3
195 #define bfd_mach_arm_3M 4
196 #define bfd_mach_arm_4 5
197 #define bfd_mach_arm_4T 6
198 #define bfd_mach_arm_5 7
199 #define bfd_mach_arm_5T 8
200 #define bfd_mach_arm_5TE 9
201 #define bfd_mach_arm_XScale 10
202 #define bfd_mach_arm_ep9312 11
203 #define bfd_mach_arm_iWMMXt 12
204 #define bfd_mach_arm_iWMMXt2 13
205 bfd_arch_ns32k, /* National Semiconductors ns32000 */
206 bfd_arch_w65, /* WDC 65816 */
207 bfd_arch_tic30, /* Texas Instruments TMS320C30 */
208 bfd_arch_v850, /* NEC V850 */
209 #define bfd_mach_v850 0
210 bfd_arch_arc, /* Argonaut RISC Core */
211 #define bfd_mach_arc_base 0
212 bfd_arch_m32r, /* Mitsubishi M32R/D */
213 #define bfd_mach_m32r 0 /* backwards compatibility */
214 bfd_arch_mn10200, /* Matsushita MN10200 */
215 bfd_arch_mn10300, /* Matsushita MN10300 */
216 bfd_arch_avr, /* AVR microcontrollers */
217 #define bfd_mach_avr1 1
218 #define bfd_mach_avr2 2
219 #define bfd_mach_avr25 25
220 #define bfd_mach_avr3 3
221 #define bfd_mach_avr31 31
222 #define bfd_mach_avr35 35
223 #define bfd_mach_avr4 4
224 #define bfd_mach_avr5 5
225 #define bfd_mach_avr51 51
226 #define bfd_mach_avr6 6
227 #define bfd_mach_avrtiny 100
228 #define bfd_mach_avrxmega1 101
229 #define bfd_mach_avrxmega2 102
230 #define bfd_mach_avrxmega3 103
231 #define bfd_mach_avrxmega4 104
232 #define bfd_mach_avrxmega5 105
233 #define bfd_mach_avrxmega6 106
234 #define bfd_mach_avrxmega7 107
235 bfd_arch_cris, /* Axis CRIS */
236 #define bfd_mach_cris_v0_v10 255
237 #define bfd_mach_cris_v32 32
238 #define bfd_mach_cris_v10_v32 1032
239 bfd_arch_microblaze, /* Xilinx MicroBlaze. */
240 bfd_arch_moxie, /* The Moxie core. */
241 bfd_arch_ia64, /* HP/Intel ia64 */
242 #define bfd_mach_ia64_elf64 64
243 #define bfd_mach_ia64_elf32 32
244 bfd_arch_rx, /* Renesas RX */
245 #define bfd_mach_rx 0x75
246 #define bfd_mach_rx_v2 0x76
247 #define bfd_mach_rx_v3 0x77
248 bfd_arch_loongarch,
249 bfd_arch_last
250 };
251 #define bfd_mach_s390_31 31
252 #define bfd_mach_s390_64 64
253
254 typedef struct symbol_cache_entry
255 {
256 const char *name;
257 union
258 {
259 PTR p;
260 bfd_vma i;
261 } udata;
262 } asymbol;
263
264 typedef int (*fprintf_function)(FILE *f, const char *fmt, ...)
265 G_GNUC_PRINTF(2, 3);
266
267 enum dis_insn_type {
268 dis_noninsn, /* Not a valid instruction */
269 dis_nonbranch, /* Not a branch instruction */
270 dis_branch, /* Unconditional branch */
271 dis_condbranch, /* Conditional branch */
272 dis_jsr, /* Jump to subroutine */
273 dis_condjsr, /* Conditional jump to subroutine */
274 dis_dref, /* Data reference instruction */
275 dis_dref2 /* Two data references in instruction */
276 };
277
278 /* This struct is passed into the instruction decoding routine,
279 and is passed back out into each callback. The various fields are used
280 for conveying information from your main routine into your callbacks,
281 for passing information into the instruction decoders (such as the
282 addresses of the callback functions), or for passing information
283 back from the instruction decoders to their callers.
284
285 It must be initialized before it is first passed; this can be done
286 by hand, or using one of the initialization macros below. */
287
288 typedef struct disassemble_info {
289 fprintf_function fprintf_func;
290 FILE *stream;
291 PTR application_data;
292
293 /* Target description. We could replace this with a pointer to the bfd,
294 but that would require one. There currently isn't any such requirement
295 so to avoid introducing one we record these explicitly. */
296 /* The bfd_flavour. This can be bfd_target_unknown_flavour. */
297 enum bfd_flavour flavour;
298 /* The bfd_arch value. */
299 enum bfd_architecture arch;
300 /* The bfd_mach value. */
301 unsigned long mach;
302 /* Endianness (for bi-endian cpus). Mono-endian cpus can ignore this. */
303 enum bfd_endian endian;
304
305 /* An array of pointers to symbols either at the location being disassembled
306 or at the start of the function being disassembled. The array is sorted
307 so that the first symbol is intended to be the one used. The others are
308 present for any misc. purposes. This is not set reliably, but if it is
309 not NULL, it is correct. */
310 asymbol **symbols;
311 /* Number of symbols in array. */
312 int num_symbols;
313
314 /* For use by the disassembler.
315 The top 16 bits are reserved for public use (and are documented here).
316 The bottom 16 bits are for the internal use of the disassembler. */
317 unsigned long flags;
318 #define INSN_HAS_RELOC 0x80000000
319 #define INSN_ARM_BE32 0x00010000
320 PTR private_data;
321
322 /* Function used to get bytes to disassemble. MEMADDR is the
323 address of the stuff to be disassembled, MYADDR is the address to
324 put the bytes in, and LENGTH is the number of bytes to read.
325 INFO is a pointer to this struct.
326 Returns an errno value or 0 for success. */
327 int (*read_memory_func)
328 (bfd_vma memaddr, bfd_byte *myaddr, int length,
329 struct disassemble_info *info);
330
331 /* Function which should be called if we get an error that we can't
332 recover from. STATUS is the errno value from read_memory_func and
333 MEMADDR is the address that we were trying to read. INFO is a
334 pointer to this struct. */
335 void (*memory_error_func)
336 (int status, bfd_vma memaddr, struct disassemble_info *info);
337
338 /* Function called to print ADDR. */
339 void (*print_address_func)
340 (bfd_vma addr, struct disassemble_info *info);
341
342 /* Function called to print an instruction. The function is architecture
343 * specific.
344 */
345 int (*print_insn)(bfd_vma addr, struct disassemble_info *info);
346
347 /* Function called to determine if there is a symbol at the given ADDR.
348 If there is, the function returns 1, otherwise it returns 0.
349 This is used by ports which support an overlay manager where
350 the overlay number is held in the top part of an address. In
351 some circumstances we want to include the overlay number in the
352 address, (normally because there is a symbol associated with
353 that address), but sometimes we want to mask out the overlay bits. */
354 int (* symbol_at_address_func)
355 (bfd_vma addr, struct disassemble_info * info);
356
357 /* These are for buffer_read_memory. */
358 const bfd_byte *buffer;
359 bfd_vma buffer_vma;
360 int buffer_length;
361
362 /* This variable may be set by the instruction decoder. It suggests
363 the number of bytes objdump should display on a single line. If
364 the instruction decoder sets this, it should always set it to
365 the same value in order to get reasonable looking output. */
366 int bytes_per_line;
367
368 /* the next two variables control the way objdump displays the raw data */
369 /* For example, if bytes_per_line is 8 and bytes_per_chunk is 4, the */
370 /* output will look like this:
371 00: 00000000 00000000
372 with the chunks displayed according to "display_endian". */
373 int bytes_per_chunk;
374 enum bfd_endian display_endian;
375
376 /* Results from instruction decoders. Not all decoders yet support
377 this information. This info is set each time an instruction is
378 decoded, and is only valid for the last such instruction.
379
380 To determine whether this decoder supports this information, set
381 insn_info_valid to 0, decode an instruction, then check it. */
382
383 char insn_info_valid; /* Branch info has been set. */
384 char branch_delay_insns; /* How many sequential insn's will run before
385 a branch takes effect. (0 = normal) */
386 char data_size; /* Size of data reference in insn, in bytes */
387 enum dis_insn_type insn_type; /* Type of instruction */
388 bfd_vma target; /* Target address of branch or dref, if known;
389 zero if unknown. */
390 bfd_vma target2; /* Second target address for dref2 */
391
392 /* Command line options specific to the target disassembler. */
393 char * disassembler_options;
394
395 /*
396 * When true instruct the disassembler it may preface the
397 * disassembly with the opcodes values if it wants to. This is
398 * mainly for the benefit of the plugin interface which doesn't want
399 * that.
400 */
401 bool show_opcodes;
402
403 /* Field intended to be used by targets in any way they deem suitable. */
404 void *target_info;
405
406 /* Options for Capstone disassembly. */
407 int cap_arch;
408 int cap_mode;
409 int cap_insn_unit;
410 int cap_insn_split;
411
412 } disassemble_info;
413
414 /* Standard disassemblers. Disassemble one instruction at the given
415 target address. Return number of bytes processed. */
416 typedef int (*disassembler_ftype) (bfd_vma, disassemble_info *);
417
418 int print_insn_tci(bfd_vma, disassemble_info*);
419 int print_insn_big_mips (bfd_vma, disassemble_info*);
420 int print_insn_little_mips (bfd_vma, disassemble_info*);
421 int print_insn_nanomips (bfd_vma, disassemble_info*);
422 int print_insn_m68k (bfd_vma, disassemble_info*);
423 int print_insn_z8001 (bfd_vma, disassemble_info*);
424 int print_insn_z8002 (bfd_vma, disassemble_info*);
425 int print_insn_h8300 (bfd_vma, disassemble_info*);
426 int print_insn_h8300h (bfd_vma, disassemble_info*);
427 int print_insn_h8300s (bfd_vma, disassemble_info*);
428 int print_insn_h8500 (bfd_vma, disassemble_info*);
429 int print_insn_arm_a64 (bfd_vma, disassemble_info*);
430 int print_insn_alpha (bfd_vma, disassemble_info*);
431 disassembler_ftype arc_get_disassembler (int, int);
432 int print_insn_sparc (bfd_vma, disassemble_info*);
433 int print_insn_big_a29k (bfd_vma, disassemble_info*);
434 int print_insn_little_a29k (bfd_vma, disassemble_info*);
435 int print_insn_i960 (bfd_vma, disassemble_info*);
436 int print_insn_sh (bfd_vma, disassemble_info*);
437 int print_insn_shl (bfd_vma, disassemble_info*);
438 int print_insn_hppa (bfd_vma, disassemble_info*);
439 int print_insn_m32r (bfd_vma, disassemble_info*);
440 int print_insn_m88k (bfd_vma, disassemble_info*);
441 int print_insn_mn10200 (bfd_vma, disassemble_info*);
442 int print_insn_mn10300 (bfd_vma, disassemble_info*);
443 int print_insn_ns32k (bfd_vma, disassemble_info*);
444 int print_insn_big_powerpc (bfd_vma, disassemble_info*);
445 int print_insn_little_powerpc (bfd_vma, disassemble_info*);
446 int print_insn_rs6000 (bfd_vma, disassemble_info*);
447 int print_insn_w65 (bfd_vma, disassemble_info*);
448 int print_insn_d10v (bfd_vma, disassemble_info*);
449 int print_insn_v850 (bfd_vma, disassemble_info*);
450 int print_insn_tic30 (bfd_vma, disassemble_info*);
451 int print_insn_crisv32 (bfd_vma, disassemble_info*);
452 int print_insn_crisv10 (bfd_vma, disassemble_info*);
453 int print_insn_microblaze (bfd_vma, disassemble_info*);
454 int print_insn_ia64 (bfd_vma, disassemble_info*);
455 int print_insn_xtensa (bfd_vma, disassemble_info*);
456 int print_insn_riscv32 (bfd_vma, disassemble_info*);
457 int print_insn_riscv64 (bfd_vma, disassemble_info*);
458 int print_insn_riscv128 (bfd_vma, disassemble_info*);
459 int print_insn_rx(bfd_vma, disassemble_info *);
460 int print_insn_hexagon(bfd_vma, disassemble_info *);
461 int print_insn_loongarch(bfd_vma, disassemble_info *);
462
463 #ifdef CONFIG_CAPSTONE
464 bool cap_disas_target(disassemble_info *info, uint64_t pc, size_t size);
465 bool cap_disas_host(disassemble_info *info, const void *code, size_t size);
466 bool cap_disas_monitor(disassemble_info *info, uint64_t pc, int count);
467 bool cap_disas_plugin(disassemble_info *info, uint64_t pc, size_t size);
468 #else
469 # define cap_disas_target(i, p, s) false
470 # define cap_disas_host(i, p, s) false
471 # define cap_disas_monitor(i, p, c) false
472 # define cap_disas_plugin(i, p, c) false
473 #endif /* CONFIG_CAPSTONE */
474
475 #ifndef ATTRIBUTE_UNUSED
476 #define ATTRIBUTE_UNUSED __attribute__((unused))
477 #endif
478
479 /* from libbfd */
480
bfd_getl64(const bfd_byte * addr)481 static inline bfd_vma bfd_getl64(const bfd_byte *addr)
482 {
483 return ldq_le_p(addr);
484 }
485
bfd_getl32(const bfd_byte * addr)486 static inline bfd_vma bfd_getl32(const bfd_byte *addr)
487 {
488 return (uint32_t)ldl_le_p(addr);
489 }
490
bfd_getl16(const bfd_byte * addr)491 static inline bfd_vma bfd_getl16(const bfd_byte *addr)
492 {
493 return lduw_le_p(addr);
494 }
495
bfd_getb32(const bfd_byte * addr)496 static inline bfd_vma bfd_getb32(const bfd_byte *addr)
497 {
498 return (uint32_t)ldl_be_p(addr);
499 }
500
bfd_getb16(const bfd_byte * addr)501 static inline bfd_vma bfd_getb16(const bfd_byte *addr)
502 {
503 return lduw_be_p(addr);
504 }
505
506 typedef bool bfd_boolean;
507
508 #endif /* DISAS_DIS_ASM_H */
509