include/dis-asm.h

   1 /* Interface between the opcode library and its callers.
   2
   3    Copyright (C) 1999-2019 Free Software Foundation, Inc.
   4
   5    This program is free software; you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 3, or (at your option)
   8    any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software
  17    Foundation, Inc., 51 Franklin Street - Fifth Floor,
  18    Boston, MA 02110-1301, USA.
  19
  20    Written by Cygnus Support, 1993.
  21
  22    The opcode library (libopcodes.a) provides instruction decoders for
  23    a large variety of instruction sets, callable with an identical
  24    interface, for making instruction-processing programs more independent
  25    of the instruction set being processed.  */
  26
  27 #ifndef DIS_ASM_H
  28 #define DIS_ASM_H
  29
  30 #ifdef __cplusplus
  31 extern "C" {
  32 #endif
  33
  34 #include <stdio.h>
  35 #include <string.h>
  36 #include "bfd.h"
  37
  38   typedef int (*fprintf_ftype) (void *, const char*, ...) ATTRIBUTE_FPTR_PRINTF_2;
  39
  40 enum dis_insn_type
  41 {
  42   dis_noninsn,                  /* Not a valid instruction.  */
  43   dis_nonbranch,                /* Not a branch instruction.  */
  44   dis_branch,                   /* Unconditional branch.  */
  45   dis_condbranch,               /* Conditional branch.  */
  46   dis_jsr,                      /* Jump to subroutine.  */
  47   dis_condjsr,                  /* Conditional jump to subroutine.  */
  48   dis_dref,                     /* Data reference instruction.  */
  49   dis_dref2                     /* Two data references in instruction.  */
  50 };
  51
  52 /* This struct is passed into the instruction decoding routine,
  53    and is passed back out into each callback.  The various fields are used
  54    for conveying information from your main routine into your callbacks,
  55    for passing information into the instruction decoders (such as the
  56    addresses of the callback functions), or for passing information
  57    back from the instruction decoders to their callers.
  58
  59    It must be initialized before it is first passed; this can be done
  60    by hand, or using one of the initialization macros below.  */
  61
  62 typedef struct disassemble_info
  63 {
  64   fprintf_ftype fprintf_func;
  65   void *stream;
  66   void *application_data;
  67
  68   /* Target description.  We could replace this with a pointer to the bfd,
  69      but that would require one.  There currently isn't any such requirement
  70      so to avoid introducing one we record these explicitly.  */
  71   /* The bfd_flavour.  This can be bfd_target_unknown_flavour.  */
  72   enum bfd_flavour flavour;
  73   /* The bfd_arch value.  */
  74   enum bfd_architecture arch;
  75   /* The bfd_mach value.  */
  76   unsigned long mach;
  77   /* Endianness (for bi-endian cpus).  Mono-endian cpus can ignore this.  */
  78   enum bfd_endian endian;
  79   /* Endianness of code, for mixed-endian situations such as ARM BE8.  */
  80   enum bfd_endian endian_code;
  81   /* An arch/mach-specific bitmask of selected instruction subsets, mainly
  82      for processors with run-time-switchable instruction sets.  The default,
  83      zero, means that there is no constraint.  CGEN-based opcodes ports
  84      may use ISA_foo masks.  */
  85   void *insn_sets;
  86
  87   /* Some targets need information about the current section to accurately
  88      display insns.  If this is NULL, the target disassembler function
  89      will have to make its best guess.  */
  90   asection *section;
  91
  92   /* An array of pointers to symbols either at the location being disassembled
  93      or at the start of the function being disassembled.  The array is sorted
  94      so that the first symbol is intended to be the one used.  The others are
  95      present for any misc. purposes.  This is not set reliably, but if it is
  96      not NULL, it is correct.  */
  97   asymbol **symbols;
  98   /* Number of symbols in array.  */
  99   int num_symbols;
 100
 101   /* Symbol table provided for targets that want to look at it.  This is
 102      used on Arm to find mapping symbols and determine Arm/Thumb code.  */
 103   asymbol **symtab;
 104   int symtab_pos;
 105   int symtab_size;
 106
 107   /* For use by the disassembler.
 108      The top 16 bits are reserved for public use (and are documented here).
 109      The bottom 16 bits are for the internal use of the disassembler.  */
 110   unsigned long flags;
 111   /* Set if the disassembler has determined that there are one or more
 112      relocations associated with the instruction being disassembled.  */
 113 #define INSN_HAS_RELOC   (1 << 31)
 114   /* Set if the user has requested the disassembly of data as well as code.  */
 115 #define DISASSEMBLE_DATA (1 << 30)
 116   /* Set if the user has specifically set the machine type encoded in the
 117      mach field of this structure.  */
 118 #define USER_SPECIFIED_MACHINE_TYPE (1 << 29)
 119
 120   /* Use internally by the target specific disassembly code.  */
 121   void *private_data;
 122
 123   /* Function used to get bytes to disassemble.  MEMADDR is the
 124      address of the stuff to be disassembled, MYADDR is the address to
 125      put the bytes in, and LENGTH is the number of bytes to read.
 126      INFO is a pointer to this struct.
 127      Returns an errno value or 0 for success.  */
 128   int (*read_memory_func)
 129     (bfd_vma memaddr, bfd_byte *myaddr, unsigned int length,
 130      struct disassemble_info *dinfo);
 131
 132   /* Function which should be called if we get an error that we can't
 133      recover from.  STATUS is the errno value from read_memory_func and
 134      MEMADDR is the address that we were trying to read.  INFO is a
 135      pointer to this struct.  */
 136   void (*memory_error_func)
 137     (int status, bfd_vma memaddr, struct disassemble_info *dinfo);
 138
 139   /* Function called to print ADDR.  */
 140   void (*print_address_func)
 141     (bfd_vma addr, struct disassemble_info *dinfo);
 142
 143   /* Function called to determine if there is a symbol at the given ADDR.
 144      If there is, the function returns 1, otherwise it returns 0.
 145      This is used by ports which support an overlay manager where
 146      the overlay number is held in the top part of an address.  In
 147      some circumstances we want to include the overlay number in the
 148      address, (normally because there is a symbol associated with
 149      that address), but sometimes we want to mask out the overlay bits.  */
 150   int (* symbol_at_address_func)
 151     (bfd_vma addr, struct disassemble_info *dinfo);
 152
 153   /* Function called to check if a SYMBOL is can be displayed to the user.
 154      This is used by some ports that want to hide special symbols when
 155      displaying debugging outout.  */
 156   bfd_boolean (* symbol_is_valid)
 157     (asymbol *, struct disassemble_info *dinfo);
 158
 159   /* These are for buffer_read_memory.  */
 160   bfd_byte *buffer;
 161   bfd_vma buffer_vma;
 162   size_t buffer_length;
 163
 164   /* This variable may be set by the instruction decoder.  It suggests
 165       the number of bytes objdump should display on a single line.  If
 166       the instruction decoder sets this, it should always set it to
 167       the same value in order to get reasonable looking output.  */
 168   int bytes_per_line;
 169
 170   /* The next two variables control the way objdump displays the raw data.  */
 171   /* For example, if bytes_per_line is 8 and bytes_per_chunk is 4, the */
 172   /* output will look like this:
 173      00:   00000000 00000000
 174      with the chunks displayed according to "display_endian". */
 175   int bytes_per_chunk;
 176   enum bfd_endian display_endian;
 177
 178   /* Number of octets per incremented target address
 179      Normally one, but some DSPs have byte sizes of 16 or 32 bits.  */
 180   unsigned int octets_per_byte;
 181
 182   /* The number of zeroes we want to see at the end of a section before we
 183      start skipping them.  */
 184   unsigned int skip_zeroes;
 185
 186   /* The number of zeroes to skip at the end of a section.  If the number
 187      of zeroes at the end is between SKIP_ZEROES_AT_END and SKIP_ZEROES,
 188      they will be disassembled.  If there are fewer than
 189      SKIP_ZEROES_AT_END, they will be skipped.  This is a heuristic
 190      attempt to avoid disassembling zeroes inserted by section
 191      alignment.  */
 192   unsigned int skip_zeroes_at_end;
 193
 194   /* Whether the disassembler always needs the relocations.  */
 195   bfd_boolean disassembler_needs_relocs;
 196
 197   /* Results from instruction decoders.  Not all decoders yet support
 198      this information.  This info is set each time an instruction is
 199      decoded, and is only valid for the last such instruction.
 200
 201      To determine whether this decoder supports this information, set
 202      insn_info_valid to 0, decode an instruction, then check it.  */
 203
 204   char insn_info_valid;         /* Branch info has been set. */
 205   char branch_delay_insns;      /* How many sequential insn's will run before
 206                                    a branch takes effect.  (0 = normal) */
 207   char data_size;               /* Size of data reference in insn, in bytes */
 208   enum dis_insn_type insn_type; /* Type of instruction */
 209   bfd_vma target;               /* Target address of branch or dref, if known;
 210                                    zero if unknown.  */
 211   bfd_vma target2;              /* Second target address for dref2 */
 212
 213   /* Command line options specific to the target disassembler.  */
 214   const char *disassembler_options;
 215
 216   /* If non-zero then try not disassemble beyond this address, even if
 217      there are values left in the buffer.  This address is the address
 218      of the nearest symbol forwards from the start of the disassembly,
 219      and it is assumed that it lies on the boundary between instructions.
 220      If an instruction spans this address then this is an error in the
 221      file being disassembled.  */
 222   bfd_vma stop_vma;
 223
 224   /* The end range of the current range being disassembled.  This is required
 225      in order to notify the disassembler when it's currently handling a
 226      different range than it was before.  This prevent unsafe optimizations when
 227      disassembling such as the way mapping symbols are found on AArch64.  */
 228   bfd_vma stop_offset;
 229
 230 } disassemble_info;
 231
 232 /* This struct is used to pass information about valid disassembler
 233    option arguments from the target to the generic GDB functions
 234    that set and display them.  */
 235
 236 typedef struct
 237 {
 238   /* Option argument name to use in descriptions.  */
 239   const char *name;
 240
 241   /* Vector of acceptable option argument values, NULL-terminated.  */
 242   const char **values;
 243 } disasm_option_arg_t;
 244
 245 /* This struct is used to pass information about valid disassembler
 246    options, their descriptions and arguments from the target to the
 247    generic GDB functions that set and display them.  Options are
 248    defined by tuples of vector entries at each index.  */
 249
 250 typedef struct
 251 {
 252   /* Vector of option names, NULL-terminated.  */
 253   const char **name;
 254
 255   /* Vector of option descriptions or NULL if none to be shown.  */
 256   const char **description;
 257
 258   /* Vector of option argument information pointers or NULL if no
 259      option accepts an argument.  NULL entries denote individual
 260      options that accept no argument.  */
 261   const disasm_option_arg_t **arg;
 262 } disasm_options_t;
 263
 264 /* This struct is used to pass information about valid disassembler
 265    options and arguments from the target to the generic GDB functions
 266    that set and display them.  */
 267
 268 typedef struct
 269 {
 270   /* Valid disassembler options.  Individual options that support
 271      an argument will refer to entries in the ARGS vector.  */
 272   disasm_options_t options;
 273
 274   /* Vector of acceptable option arguments, NULL-terminated.  This
 275      collects all possible option argument choices, some of which
 276      may be shared by different options from the OPTIONS member.  */
 277   disasm_option_arg_t *args;
 278 } disasm_options_and_args_t;
 279 \f
 280 /* Standard disassemblers.  Disassemble one instruction at the given
 281    target address.  Return number of octets processed.  */
 282 typedef int (*disassembler_ftype) (bfd_vma, disassemble_info *);
 283
 284 /* Disassemblers used out side of opcodes library.  */
 285 extern int print_insn_m32c              (bfd_vma, disassemble_info *);
 286 extern int print_insn_mep               (bfd_vma, disassemble_info *);
 287 extern int print_insn_s12z              (bfd_vma, disassemble_info *);
 288 extern int print_insn_sh                (bfd_vma, disassemble_info *);
 289 extern int print_insn_sparc             (bfd_vma, disassemble_info *);
 290 extern int print_insn_rx                (bfd_vma, disassemble_info *);
 291 extern int print_insn_rl78              (bfd_vma, disassemble_info *);
 292 extern int print_insn_rl78_g10          (bfd_vma, disassemble_info *);
 293 extern int print_insn_rl78_g13          (bfd_vma, disassemble_info *);
 294 extern int print_insn_rl78_g14          (bfd_vma, disassemble_info *);
 295
 296 extern disassembler_ftype arc_get_disassembler (bfd *);
 297 extern disassembler_ftype cris_get_disassembler (bfd *);
 298
 299 extern void print_aarch64_disassembler_options (FILE *);
 300 extern void print_i386_disassembler_options (FILE *);
 301 extern void print_mips_disassembler_options (FILE *);
 302 extern void print_nfp_disassembler_options (FILE *);
 303 extern void print_ppc_disassembler_options (FILE *);
 304 extern void print_riscv_disassembler_options (FILE *);
 305 extern void print_arm_disassembler_options (FILE *);
 306 extern void print_arc_disassembler_options (FILE *);
 307 extern void print_s390_disassembler_options (FILE *);
 308 extern void print_wasm32_disassembler_options (FILE *);
 309 extern bfd_boolean aarch64_symbol_is_valid (asymbol *, struct disassemble_info *);
 310 extern bfd_boolean arm_symbol_is_valid (asymbol *, struct disassemble_info *);
 311 extern bfd_boolean csky_symbol_is_valid (asymbol *, struct disassemble_info *);
 312 extern bfd_boolean riscv_symbol_is_valid (asymbol *, struct disassemble_info *);
 313 extern void disassemble_init_powerpc (struct disassemble_info *);
 314 extern void disassemble_init_s390 (struct disassemble_info *);
 315 extern void disassemble_init_wasm32 (struct disassemble_info *);
 316 extern void disassemble_init_nds32 (struct disassemble_info *);
 317 extern const disasm_options_and_args_t *disassembler_options_arm (void);
 318 extern const disasm_options_and_args_t *disassembler_options_mips (void);
 319 extern const disasm_options_and_args_t *disassembler_options_powerpc (void);
 320 extern const disasm_options_and_args_t *disassembler_options_s390 (void);
 321
 322 /* Fetch the disassembler for a given architecture ARC, endianess (big
 323    endian if BIG is true), bfd_mach value MACH, and ABFD, if that support
 324    is available.  ABFD may be NULL.  */
 325 extern disassembler_ftype disassembler (enum bfd_architecture arc,
 326                                         bfd_boolean big, unsigned long mach,
 327                                         bfd *abfd);
 328
 329 /* Amend the disassemble_info structure as necessary for the target architecture.
 330    Should only be called after initialising the info->arch field.  */
 331 extern void disassemble_init_for_target (struct disassemble_info * dinfo);
 332
 333 /* Document any target specific options available from the disassembler.  */
 334 extern void disassembler_usage (FILE *);
 335
 336 /* Remove whitespace and consecutive commas.  */
 337 extern char *remove_whitespace_and_extra_commas (char *);
 338
 339 /* Like STRCMP, but treat ',' the same as '\0' so that we match
 340    strings like "foobar" against "foobar,xxyyzz,...".  */
 341 extern int disassembler_options_cmp (const char *, const char *);
 342
 343 /* A helper function for FOR_EACH_DISASSEMBLER_OPTION.  */
 344 static inline const char *
 345 next_disassembler_option (const char *options)
 346 {
 347   const char *opt = strchr (options, ',');
 348   if (opt != NULL)
 349     opt++;
 350   return opt;
 351 }
 352
 353 /* A macro for iterating over each comma separated option in OPTIONS.  */
 354 #define FOR_EACH_DISASSEMBLER_OPTION(OPT, OPTIONS) \
 355   for ((OPT) = (OPTIONS); \
 356        (OPT) != NULL; \
 357        (OPT) = next_disassembler_option (OPT))
 358
 359 \f
 360 /* This block of definitions is for particular callers who read instructions
 361    into a buffer before calling the instruction decoder.  */
 362
 363 /* Here is a function which callers may wish to use for read_memory_func.
 364    It gets bytes from a buffer.  */
 365 extern int buffer_read_memory
 366   (bfd_vma, bfd_byte *, unsigned int, struct disassemble_info *);
 367
 368 /* This function goes with buffer_read_memory.
 369    It prints a message using info->fprintf_func and info->stream.  */
 370 extern void perror_memory (int, bfd_vma, struct disassemble_info *);
 371
 372
 373 /* Just print the address in hex.  This is included for completeness even
 374    though both GDB and objdump provide their own (to print symbolic
 375    addresses).  */
 376 extern void generic_print_address
 377   (bfd_vma, struct disassemble_info *);
 378
 379 /* Always true.  */
 380 extern int generic_symbol_at_address
 381   (bfd_vma, struct disassemble_info *);
 382
 383 /* Also always true.  */
 384 extern bfd_boolean generic_symbol_is_valid
 385   (asymbol *, struct disassemble_info *);
 386
 387 /* Method to initialize a disassemble_info struct.  This should be
 388    called by all applications creating such a struct.  */
 389 extern void init_disassemble_info (struct disassemble_info *dinfo, void *stream,
 390                                    fprintf_ftype fprintf_func);
 391
 392 /* For compatibility with existing code.  */
 393 #define INIT_DISASSEMBLE_INFO(INFO, STREAM, FPRINTF_FUNC) \
 394   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 395 #define INIT_DISASSEMBLE_INFO_NO_ARCH(INFO, STREAM, FPRINTF_FUNC) \
 396   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 397
 398
 399 #ifdef __cplusplus
 400 }
 401 #endif
 402
 403 #endif /* ! defined (DIS_ASM_H) */