include/dis-asm.h

   1 /* Interface between the opcode library and its callers.
   2
   3    Copyright (C) 1999-2017 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 "bfd.h"
  36
  37   typedef int (*fprintf_ftype) (void *, const char*, ...) ATTRIBUTE_FPTR_PRINTF_2;
  38
  39 enum dis_insn_type
  40 {
  41   dis_noninsn,                  /* Not a valid instruction.  */
  42   dis_nonbranch,                /* Not a branch instruction.  */
  43   dis_branch,                   /* Unconditional branch.  */
  44   dis_condbranch,               /* Conditional branch.  */
  45   dis_jsr,                      /* Jump to subroutine.  */
  46   dis_condjsr,                  /* Conditional jump to subroutine.  */
  47   dis_dref,                     /* Data reference instruction.  */
  48   dis_dref2                     /* Two data references in instruction.  */
  49 };
  50
  51 /* This struct is passed into the instruction decoding routine,
  52    and is passed back out into each callback.  The various fields are used
  53    for conveying information from your main routine into your callbacks,
  54    for passing information into the instruction decoders (such as the
  55    addresses of the callback functions), or for passing information
  56    back from the instruction decoders to their callers.
  57
  58    It must be initialized before it is first passed; this can be done
  59    by hand, or using one of the initialization macros below.  */
  60
  61 typedef struct disassemble_info
  62 {
  63   fprintf_ftype fprintf_func;
  64   void *stream;
  65   void *application_data;
  66
  67   /* Target description.  We could replace this with a pointer to the bfd,
  68      but that would require one.  There currently isn't any such requirement
  69      so to avoid introducing one we record these explicitly.  */
  70   /* The bfd_flavour.  This can be bfd_target_unknown_flavour.  */
  71   enum bfd_flavour flavour;
  72   /* The bfd_arch value.  */
  73   enum bfd_architecture arch;
  74   /* The bfd_mach value.  */
  75   unsigned long mach;
  76   /* Endianness (for bi-endian cpus).  Mono-endian cpus can ignore this.  */
  77   enum bfd_endian endian;
  78   /* Endianness of code, for mixed-endian situations such as ARM BE8.  */
  79   enum bfd_endian endian_code;
  80   /* An arch/mach-specific bitmask of selected instruction subsets, mainly
  81      for processors with run-time-switchable instruction sets.  The default,
  82      zero, means that there is no constraint.  CGEN-based opcodes ports
  83      may use ISA_foo masks.  */
  84   void *insn_sets;
  85
  86   /* Some targets need information about the current section to accurately
  87      display insns.  If this is NULL, the target disassembler function
  88      will have to make its best guess.  */
  89   asection *section;
  90
  91   /* An array of pointers to symbols either at the location being disassembled
  92      or at the start of the function being disassembled.  The array is sorted
  93      so that the first symbol is intended to be the one used.  The others are
  94      present for any misc. purposes.  This is not set reliably, but if it is
  95      not NULL, it is correct.  */
  96   asymbol **symbols;
  97   /* Number of symbols in array.  */
  98   int num_symbols;
  99
 100   /* Symbol table provided for targets that want to look at it.  This is
 101      used on Arm to find mapping symbols and determine Arm/Thumb code.  */
 102   asymbol **symtab;
 103   int symtab_pos;
 104   int symtab_size;
 105
 106   /* For use by the disassembler.
 107      The top 16 bits are reserved for public use (and are documented here).
 108      The bottom 16 bits are for the internal use of the disassembler.  */
 109   unsigned long flags;
 110   /* Set if the disassembler has determined that there are one or more
 111      relocations associated with the instruction being disassembled.  */
 112 #define INSN_HAS_RELOC   (1 << 31)
 113   /* Set if the user has requested the disassembly of data as well as code.  */
 114 #define DISASSEMBLE_DATA (1 << 30)
 115   /* Set if the user has specifically set the machine type encoded in the
 116      mach field of this structure.  */
 117 #define USER_SPECIFIED_MACHINE_TYPE (1 << 29)
 118
 119   /* Use internally by the target specific disassembly code.  */
 120   void *private_data;
 121
 122   /* Function used to get bytes to disassemble.  MEMADDR is the
 123      address of the stuff to be disassembled, MYADDR is the address to
 124      put the bytes in, and LENGTH is the number of bytes to read.
 125      INFO is a pointer to this struct.
 126      Returns an errno value or 0 for success.  */
 127   int (*read_memory_func)
 128     (bfd_vma memaddr, bfd_byte *myaddr, unsigned int length,
 129      struct disassemble_info *dinfo);
 130
 131   /* Function which should be called if we get an error that we can't
 132      recover from.  STATUS is the errno value from read_memory_func and
 133      MEMADDR is the address that we were trying to read.  INFO is a
 134      pointer to this struct.  */
 135   void (*memory_error_func)
 136     (int status, bfd_vma memaddr, struct disassemble_info *dinfo);
 137
 138   /* Function called to print ADDR.  */
 139   void (*print_address_func)
 140     (bfd_vma addr, struct disassemble_info *dinfo);
 141
 142   /* Function called to determine if there is a symbol at the given ADDR.
 143      If there is, the function returns 1, otherwise it returns 0.
 144      This is used by ports which support an overlay manager where
 145      the overlay number is held in the top part of an address.  In
 146      some circumstances we want to include the overlay number in the
 147      address, (normally because there is a symbol associated with
 148      that address), but sometimes we want to mask out the overlay bits.  */
 149   int (* symbol_at_address_func)
 150     (bfd_vma addr, struct disassemble_info *dinfo);
 151
 152   /* Function called to check if a SYMBOL is can be displayed to the user.
 153      This is used by some ports that want to hide special symbols when
 154      displaying debugging outout.  */
 155   bfd_boolean (* symbol_is_valid)
 156     (asymbol *, struct disassemble_info *dinfo);
 157
 158   /* These are for buffer_read_memory.  */
 159   bfd_byte *buffer;
 160   bfd_vma buffer_vma;
 161   unsigned int buffer_length;
 162
 163   /* This variable may be set by the instruction decoder.  It suggests
 164       the number of bytes objdump should display on a single line.  If
 165       the instruction decoder sets this, it should always set it to
 166       the same value in order to get reasonable looking output.  */
 167   int bytes_per_line;
 168
 169   /* The next two variables control the way objdump displays the raw data.  */
 170   /* For example, if bytes_per_line is 8 and bytes_per_chunk is 4, the */
 171   /* output will look like this:
 172      00:   00000000 00000000
 173      with the chunks displayed according to "display_endian". */
 174   int bytes_per_chunk;
 175   enum bfd_endian display_endian;
 176
 177   /* Number of octets per incremented target address
 178      Normally one, but some DSPs have byte sizes of 16 or 32 bits.  */
 179   unsigned int octets_per_byte;
 180
 181   /* The number of zeroes we want to see at the end of a section before we
 182      start skipping them.  */
 183   unsigned int skip_zeroes;
 184
 185   /* The number of zeroes to skip at the end of a section.  If the number
 186      of zeroes at the end is between SKIP_ZEROES_AT_END and SKIP_ZEROES,
 187      they will be disassembled.  If there are fewer than
 188      SKIP_ZEROES_AT_END, they will be skipped.  This is a heuristic
 189      attempt to avoid disassembling zeroes inserted by section
 190      alignment.  */
 191   unsigned int skip_zeroes_at_end;
 192
 193   /* Whether the disassembler always needs the relocations.  */
 194   bfd_boolean disassembler_needs_relocs;
 195
 196   /* Results from instruction decoders.  Not all decoders yet support
 197      this information.  This info is set each time an instruction is
 198      decoded, and is only valid for the last such instruction.
 199
 200      To determine whether this decoder supports this information, set
 201      insn_info_valid to 0, decode an instruction, then check it.  */
 202
 203   char insn_info_valid;         /* Branch info has been set. */
 204   char branch_delay_insns;      /* How many sequential insn's will run before
 205                                    a branch takes effect.  (0 = normal) */
 206   char data_size;               /* Size of data reference in insn, in bytes */
 207   enum dis_insn_type insn_type; /* Type of instruction */
 208   bfd_vma target;               /* Target address of branch or dref, if known;
 209                                    zero if unknown.  */
 210   bfd_vma target2;              /* Second target address for dref2 */
 211
 212   /* Command line options specific to the target disassembler.  */
 213   const char *disassembler_options;
 214
 215   /* If non-zero then try not disassemble beyond this address, even if
 216      there are values left in the buffer.  This address is the address
 217      of the nearest symbol forwards from the start of the disassembly,
 218      and it is assumed that it lies on the boundary between instructions.
 219      If an instruction spans this address then this is an error in the
 220      file being disassembled.  */
 221   bfd_vma stop_vma;
 222
 223 } disassemble_info;
 224
 225 /* This struct is used to pass information about valid disassembler options
 226    and their descriptions from the target to the generic GDB functions that
 227    set and display them.  */
 228
 229 typedef struct
 230 {
 231   const char **name;
 232   const char **description;
 233 } disasm_options_t;
 234
 235 \f
 236 /* Standard disassemblers.  Disassemble one instruction at the given
 237    target address.  Return number of octets processed.  */
 238 typedef int (*disassembler_ftype) (bfd_vma, disassemble_info *);
 239
 240 /* Disassemblers used out side of opcodes library.  */
 241 extern int print_insn_aarch64           (bfd_vma, disassemble_info *);
 242 extern int print_insn_big_arm           (bfd_vma, disassemble_info *);
 243 extern int print_insn_big_mips          (bfd_vma, disassemble_info *);
 244 extern int print_insn_i386              (bfd_vma, disassemble_info *);
 245 extern int print_insn_ia64              (bfd_vma, disassemble_info *);
 246 extern int print_insn_little_arm        (bfd_vma, disassemble_info *);
 247 extern int print_insn_little_mips       (bfd_vma, disassemble_info *);
 248 extern int print_insn_m32c              (bfd_vma, disassemble_info *);
 249 extern int print_insn_mep               (bfd_vma, disassemble_info *);
 250 extern int print_insn_sh                (bfd_vma, disassemble_info *);
 251 extern int print_insn_sh64x_media       (bfd_vma, disassemble_info *);
 252 extern int print_insn_sparc             (bfd_vma, disassemble_info *);
 253 extern int print_insn_spu               (bfd_vma, disassemble_info *);
 254 extern int print_insn_rx                (bfd_vma, disassemble_info *);
 255 extern int print_insn_rl78              (bfd_vma, disassemble_info *);
 256 extern int print_insn_rl78_g10          (bfd_vma, disassemble_info *);
 257 extern int print_insn_rl78_g13          (bfd_vma, disassemble_info *);
 258 extern int print_insn_rl78_g14          (bfd_vma, disassemble_info *);
 259
 260 extern disassembler_ftype arc_get_disassembler (bfd *);
 261 extern disassembler_ftype cris_get_disassembler (bfd *);
 262
 263 extern void print_aarch64_disassembler_options (FILE *);
 264 extern void print_i386_disassembler_options (FILE *);
 265 extern void print_mips_disassembler_options (FILE *);
 266 extern void print_ppc_disassembler_options (FILE *);
 267 extern void print_riscv_disassembler_options (FILE *);
 268 extern void print_arm_disassembler_options (FILE *);
 269 extern void print_arc_disassembler_options (FILE *);
 270 extern void print_s390_disassembler_options (FILE *);
 271 extern void print_wasm32_disassembler_options (FILE *);
 272 extern bfd_boolean aarch64_symbol_is_valid (asymbol *, struct disassemble_info *);
 273 extern bfd_boolean arm_symbol_is_valid (asymbol *, struct disassemble_info *);
 274 extern void disassemble_init_powerpc (struct disassemble_info *);
 275 extern void disassemble_init_s390 (struct disassemble_info *);
 276 extern void disassemble_init_wasm32 (struct disassemble_info *);
 277 extern const disasm_options_t *disassembler_options_powerpc (void);
 278 extern const disasm_options_t *disassembler_options_arm (void);
 279 extern const disasm_options_t *disassembler_options_s390 (void);
 280
 281 /* Fetch the disassembler for a given architecture ARC, endianess (big
 282    endian if BIG is true), bfd_mach value MACH, and ABFD, if that support
 283    is available.  ABFD may be NULL.  */
 284 extern disassembler_ftype disassembler (enum bfd_architecture arc,
 285                                         bfd_boolean big, unsigned long mach,
 286                                         bfd *abfd);
 287
 288 /* Amend the disassemble_info structure as necessary for the target architecture.
 289    Should only be called after initialising the info->arch field.  */
 290 extern void disassemble_init_for_target (struct disassemble_info * dinfo);
 291
 292 /* Document any target specific options available from the disassembler.  */
 293 extern void disassembler_usage (FILE *);
 294
 295 /* Remove whitespace and consecutive commas.  */
 296 extern char *remove_whitespace_and_extra_commas (char *);
 297
 298 /* Like STRCMP, but treat ',' the same as '\0' so that we match
 299    strings like "foobar" against "foobar,xxyyzz,...".  */
 300 extern int disassembler_options_cmp (const char *, const char *);
 301
 302 /* A helper function for FOR_EACH_DISASSEMBLER_OPTION.  */
 303 static inline const char *
 304 next_disassembler_option (const char *options)
 305 {
 306   const char *opt = strchr (options, ',');
 307   if (opt != NULL)
 308     opt++;
 309   return opt;
 310 }
 311
 312 /* A macro for iterating over each comma separated option in OPTIONS.  */
 313 #define FOR_EACH_DISASSEMBLER_OPTION(OPT, OPTIONS) \
 314   for ((OPT) = (OPTIONS); \
 315        (OPT) != NULL; \
 316        (OPT) = next_disassembler_option (OPT))
 317
 318 \f
 319 /* This block of definitions is for particular callers who read instructions
 320    into a buffer before calling the instruction decoder.  */
 321
 322 /* Here is a function which callers may wish to use for read_memory_func.
 323    It gets bytes from a buffer.  */
 324 extern int buffer_read_memory
 325   (bfd_vma, bfd_byte *, unsigned int, struct disassemble_info *);
 326
 327 /* This function goes with buffer_read_memory.
 328    It prints a message using info->fprintf_func and info->stream.  */
 329 extern void perror_memory (int, bfd_vma, struct disassemble_info *);
 330
 331
 332 /* Just print the address in hex.  This is included for completeness even
 333    though both GDB and objdump provide their own (to print symbolic
 334    addresses).  */
 335 extern void generic_print_address
 336   (bfd_vma, struct disassemble_info *);
 337
 338 /* Always true.  */
 339 extern int generic_symbol_at_address
 340   (bfd_vma, struct disassemble_info *);
 341
 342 /* Also always true.  */
 343 extern bfd_boolean generic_symbol_is_valid
 344   (asymbol *, struct disassemble_info *);
 345
 346 /* Method to initialize a disassemble_info struct.  This should be
 347    called by all applications creating such a struct.  */
 348 extern void init_disassemble_info (struct disassemble_info *dinfo, void *stream,
 349                                    fprintf_ftype fprintf_func);
 350
 351 /* For compatibility with existing code.  */
 352 #define INIT_DISASSEMBLE_INFO(INFO, STREAM, FPRINTF_FUNC) \
 353   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 354 #define INIT_DISASSEMBLE_INFO_NO_ARCH(INFO, STREAM, FPRINTF_FUNC) \
 355   init_disassemble_info (&(INFO), (STREAM), (fprintf_ftype) (FPRINTF_FUNC))
 356
 357
 358 #ifdef __cplusplus
 359 }
 360 #endif
 361
 362 #endif /* ! defined (DIS_ASM_H) */