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   /* Set if the user has requested wide output.  */
 120 #define WIDE_OUTPUT (1 << 28)
 121
 122   /* Use internally by the target specific disassembly code.  */
 123   void *private_data;
 124
 125   /* Function used to get bytes to disassemble.  MEMADDR is the
 126      address of the stuff to be disassembled, MYADDR is the address to
 127      put the bytes in, and LENGTH is the number of bytes to read.
 128      INFO is a pointer to this struct.
 129      Returns an errno value or 0 for success.  */
 130   int (*read_memory_func)
 131     (bfd_vma memaddr, bfd_byte *myaddr, unsigned int length,
 132      struct disassemble_info *dinfo);
 133
 134   /* Function which should be called if we get an error that we can't
 135      recover from.  STATUS is the errno value from read_memory_func and
 136      MEMADDR is the address that we were trying to read.  INFO is a
 137      pointer to this struct.  */
 138   void (*memory_error_func)
 139     (int status, bfd_vma memaddr, struct disassemble_info *dinfo);
 140
 141   /* Function called to print ADDR.  */
 142   void (*print_address_func)
 143     (bfd_vma addr, struct disassemble_info *dinfo);
 144
 145   /* Function called to determine if there is a symbol at the given ADDR.
 146      If there is, the function returns 1, otherwise it returns 0.
 147      This is used by ports which support an overlay manager where
 148      the overlay number is held in the top part of an address.  In
 149      some circumstances we want to include the overlay number in the
 150      address, (normally because there is a symbol associated with
 151      that address), but sometimes we want to mask out the overlay bits.  */
 152   int (* symbol_at_address_func)
 153     (bfd_vma addr, struct disassemble_info *dinfo);
 154
 155   /* Function called to check if a SYMBOL is can be displayed to the user.
 156      This is used by some ports that want to hide special symbols when
 157      displaying debugging outout.  */
 158   bfd_boolean (* symbol_is_valid)
 159     (asymbol *, struct disassemble_info *dinfo);
 160
 161   /* These are for buffer_read_memory.  */
 162   bfd_byte *buffer;
 163   bfd_vma buffer_vma;
 164   size_t buffer_length;
 165
 166   /* This variable may be set by the instruction decoder.  It suggests
 167       the number of bytes objdump should display on a single line.  If
 168       the instruction decoder sets this, it should always set it to
 169       the same value in order to get reasonable looking output.  */
 170   int bytes_per_line;
 171
 172   /* The next two variables control the way objdump displays the raw data.  */
 173   /* For example, if bytes_per_line is 8 and bytes_per_chunk is 4, the */
 174   /* output will look like this:
 175      00:   00000000 00000000
 176      with the chunks displayed according to "display_endian". */
 177   int bytes_per_chunk;
 178   enum bfd_endian display_endian;
 179
 180   /* Number of octets per incremented target address
 181      Normally one, but some DSPs have byte sizes of 16 or 32 bits.  */
 182   unsigned int octets_per_byte;
 183
 184   /* The number of zeroes we want to see at the end of a section before we
 185      start skipping them.  */
 186   unsigned int skip_zeroes;
 187
 188   /* The number of zeroes to skip at the end of a section.  If the number
 189      of zeroes at the end is between SKIP_ZEROES_AT_END and SKIP_ZEROES,
 190      they will be disassembled.  If there are fewer than
 191      SKIP_ZEROES_AT_END, they will be skipped.  This is a heuristic
 192      attempt to avoid disassembling zeroes inserted by section
 193      alignment.  */
 194   unsigned int skip_zeroes_at_end;
 195
 196   /* Whether the disassembler always needs the relocations.  */
 197   bfd_boolean disassembler_needs_relocs;
 198
 199   /* Results from instruction decoders.  Not all decoders yet support
 200      this information.  This info is set each time an instruction is
 201      decoded, and is only valid for the last such instruction.
 202
 203      To determine whether this decoder supports this information, set
 204      insn_info_valid to 0, decode an instruction, then check it.  */
 205
 206   char insn_info_valid;         /* Branch info has been set. */
 207   char branch_delay_insns;      /* How many sequential insn's will run before
 208                                    a branch takes effect.  (0 = normal) */
 209   char data_size;               /* Size of data reference in insn, in bytes */
 210   enum dis_insn_type insn_type; /* Type of instruction */
 211   bfd_vma target;               /* Target address of branch or dref, if known;
 212                                    zero if unknown.  */
 213   bfd_vma target2;              /* Second target address for dref2 */
 214
 215   /* Command line options specific to the target disassembler.  */
 216   const char *disassembler_options;
 217
 218   /* If non-zero then try not disassemble beyond this address, even if
 219      there are values left in the buffer.  This address is the address
 220      of the nearest symbol forwards from the start of the disassembly,
 221      and it is assumed that it lies on the boundary between instructions.
 222      If an instruction spans this address then this is an error in the
 223      file being disassembled.  */
 224   bfd_vma stop_vma;
 225
 226   /* The end range of the current range being disassembled.  This is required
 227      in order to notify the disassembler when it's currently handling a
 228      different range than it was before.  This prevent unsafe optimizations when
 229      disassembling such as the way mapping symbols are found on AArch64.  */
 230   bfd_vma stop_offset;
 231
 232 } disassemble_info;
 233
 234 /* This struct is used to pass information about valid disassembler
 235    option arguments from the target to the generic GDB functions
 236    that set and display them.  */
 237
 238 typedef struct
 239 {
 240   /* Option argument name to use in descriptions.  */
 241   const char *name;
 242
 243   /* Vector of acceptable option argument values, NULL-terminated.  */
 244   const char **values;
 245 } disasm_option_arg_t;
 246
 247 /* This struct is used to pass information about valid disassembler
 248    options, their descriptions and arguments from the target to the
 249    generic GDB functions that set and display them.  Options are
 250    defined by tuples of vector entries at each index.  */
 251
 252 typedef struct
 253 {
 254   /* Vector of option names, NULL-terminated.  */
 255   const char **name;
 256
 257   /* Vector of option descriptions or NULL if none to be shown.  */
 258   const char **description;
 259
 260   /* Vector of option argument information pointers or NULL if no
 261      option accepts an argument.  NULL entries denote individual
 262      options that accept no argument.  */
 263   const disasm_option_arg_t **arg;
 264 } disasm_options_t;
 265
 266 /* This struct is used to pass information about valid disassembler
 267    options and arguments from the target to the generic GDB functions
 268    that set and display them.  */
 269
 270 typedef struct
 271 {
 272   /* Valid disassembler options.  Individual options that support
 273      an argument will refer to entries in the ARGS vector.  */
 274   disasm_options_t options;
 275
 276   /* Vector of acceptable option arguments, NULL-terminated.  This
 277      collects all possible option argument choices, some of which
 278      may be shared by different options from the OPTIONS member.  */
 279   disasm_option_arg_t *args;
 280 } disasm_options_and_args_t;
 281 \f
 282 /* Standard disassemblers.  Disassemble one instruction at the given
 283    target address.  Return number of octets processed.  */
 284 typedef int (*disassembler_ftype) (bfd_vma, disassemble_info *);
 285
 286 /* Disassemblers used out side of opcodes library.  */
 287 extern int print_insn_m32c              (bfd_vma, disassemble_info *);
 288 extern int print_insn_mep               (bfd_vma, disassemble_info *);
 289 extern int print_insn_s12z              (bfd_vma, disassemble_info *);
 290 extern int print_insn_sh                (bfd_vma, disassemble_info *);
 291 extern int print_insn_sparc             (bfd_vma, disassemble_info *);
 292 extern int print_insn_rx                (bfd_vma, disassemble_info *);
 293 extern int print_insn_rl78              (bfd_vma, disassemble_info *);
 294 extern int print_insn_rl78_g10          (bfd_vma, disassemble_info *);
 295 extern int print_insn_rl78_g13          (bfd_vma, disassemble_info *);
 296 extern int print_insn_rl78_g14          (bfd_vma, disassemble_info *);
 297
 298 extern disassembler_ftype arc_get_disassembler (bfd *);
 299 extern disassembler_ftype cris_get_disassembler (bfd *);
 300
 301 extern void print_aarch64_disassembler_options (FILE *);
 302 extern void print_i386_disassembler_options (FILE *);
 303 extern void print_mips_disassembler_options (FILE *);
 304 extern void print_nfp_disassembler_options (FILE *);
 305 extern void print_ppc_disassembler_options (FILE *);
 306 extern void print_riscv_disassembler_options (FILE *);
 307 extern void print_arm_disassembler_options (FILE *);
 308 extern void print_arc_disassembler_options (FILE *);
 309 extern void print_s390_disassembler_options (FILE *);
 310 extern void print_wasm32_disassembler_options (FILE *);
 311 extern bfd_boolean aarch64_symbol_is_valid (asymbol *, struct disassemble_info *);
 312 extern bfd_boolean arm_symbol_is_valid (asymbol *, struct disassemble_info *);
 313 extern bfd_boolean csky_symbol_is_valid (asymbol *, struct disassemble_info *);
 314 extern bfd_boolean riscv_symbol_is_valid (asymbol *, struct disassemble_info *);
 315 extern void disassemble_init_powerpc (struct disassemble_info *);
 316 extern void disassemble_init_s390 (struct disassemble_info *);
 317 extern void disassemble_init_wasm32 (struct disassemble_info *);
 318 extern void disassemble_init_nds32 (struct disassemble_info *);
 319 extern const disasm_options_and_args_t *disassembler_options_arm (void);
 320 extern const disasm_options_and_args_t *disassembler_options_mips (void);
 321 extern const disasm_options_and_args_t *disassembler_options_powerpc (void);
 322 extern const disasm_options_and_args_t *disassembler_options_s390 (void);
 323
 324 /* Fetch the disassembler for a given architecture ARC, endianess (big
 325    endian if BIG is true), bfd_mach value MACH, and ABFD, if that support
 326    is available.  ABFD may be NULL.  */
 327 extern disassembler_ftype disassembler (enum bfd_architecture arc,
 328                                         bfd_boolean big, unsigned long mach,
 329                                         bfd *abfd);
 330
 331 /* Amend the disassemble_info structure as necessary for the target architecture.
 332    Should only be called after initialising the info->arch field.  */
 333 extern void disassemble_init_for_target (struct disassemble_info * dinfo);
 334
 335 /* Document any target specific options available from the disassembler.  */
 336 extern void disassembler_usage (FILE *);
 337
 338 /* Remove whitespace and consecutive commas.  */
 339 extern char *remove_whitespace_and_extra_commas (char *);
 340
 341 /* Like STRCMP, but treat ',' the same as '\0' so that we match
 342    strings like "foobar" against "foobar,xxyyzz,...".  */
 343 extern int disassembler_options_cmp (const char *, const char *);
 344
 345 /* A helper function for FOR_EACH_DISASSEMBLER_OPTION.  */
 346 static inline const char *
 347 next_disassembler_option (const char *options)
 348 {
 349   const char *opt = strchr (options, ',');
 350   if (opt != NULL)
 351     opt++;
 352   return opt;
 353 }
 354
 355 /* A macro for iterating over each comma separated option in OPTIONS.  */
 356 #define FOR_EACH_DISASSEMBLER_OPTION(OPT, OPTIONS) \
 357   for ((OPT) = (OPTIONS); \
 358        (OPT) != NULL; \
 359        (OPT) = next_disassembler_option (OPT))
 360
 361 \f
 362 /* This block of definitions is for particular callers who read instructions
 363    into a buffer before calling the instruction decoder.  */
 364
 365 /* Here is a function which callers may wish to use for read_memory_func.
 366    It gets bytes from a buffer.  */
 367 extern int buffer_read_memory
 368   (bfd_vma, bfd_byte *, unsigned int, struct disassemble_info *);
 369
 370 /* This function goes with buffer_read_memory.
 371    It prints a message using info->fprintf_func and info->stream.  */
 372 extern void perror_memory (int, bfd_vma, struct disassemble_info *);
 373
 374
 375 /* Just print the address in hex.  This is included for completeness even
 376    though both GDB and objdump provide their own (to print symbolic
 377    addresses).  */
 378 extern void generic_print_address
 379   (bfd_vma, struct disassemble_info *);
 380
 381 /* Always true.  */
 382 extern int generic_symbol_at_address
 383   (bfd_vma, struct disassemble_info *);
 384
 385 /* Also always true.  */
 386 extern bfd_boolean generic_symbol_is_valid
 387   (asymbol *, struct disassemble_info *);
 388
 389 /* Method to initialize a disassemble_info struct.  This should be
 390    called by all applications creating such a struct.  */
 391 extern void init_disassemble_info (struct disassemble_info *dinfo, void *stream,
 392                                    fprintf_ftype fprintf_func);
 393
 394 /* For compatibility with existing code.  */
 395 #define INIT_DISASSEMBLE_INFO(INFO, STREAM, FPRINTF_FUNC) \
 396   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 397 #define INIT_DISASSEMBLE_INFO_NO_ARCH(INFO, STREAM, FPRINTF_FUNC) \
 398   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 399
 400
 401 #ifdef __cplusplus
 402 }
 403 #endif
 404
 405 #endif /* ! defined (DIS_ASM_H) */