include/dis-asm.h

   1 /* Interface between the opcode library and its callers.
   2
   3    Copyright (C) 1999-2020 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
  82   /* Some targets need information about the current section to accurately
  83      display insns.  If this is NULL, the target disassembler function
  84      will have to make its best guess.  */
  85   asection *section;
  86
  87   /* An array of pointers to symbols either at the location being disassembled
  88      or at the start of the function being disassembled.  The array is sorted
  89      so that the first symbol is intended to be the one used.  The others are
  90      present for any misc. purposes.  This is not set reliably, but if it is
  91      not NULL, it is correct.  */
  92   asymbol **symbols;
  93   /* Number of symbols in array.  */
  94   int num_symbols;
  95
  96   /* Symbol table provided for targets that want to look at it.  This is
  97      used on Arm to find mapping symbols and determine Arm/Thumb code.  */
  98   asymbol **symtab;
  99   int symtab_pos;
 100   int symtab_size;
 101
 102   /* For use by the disassembler.
 103      The top 16 bits are reserved for public use (and are documented here).
 104      The bottom 16 bits are for the internal use of the disassembler.  */
 105   unsigned long flags;
 106   /* Set if the disassembler has determined that there are one or more
 107      relocations associated with the instruction being disassembled.  */
 108 #define INSN_HAS_RELOC   (1u << 31)
 109   /* Set if the user has requested the disassembly of data as well as code.  */
 110 #define DISASSEMBLE_DATA (1u << 30)
 111   /* Set if the user has specifically set the machine type encoded in the
 112      mach field of this structure.  */
 113 #define USER_SPECIFIED_MACHINE_TYPE (1u << 29)
 114   /* Set if the user has requested wide output.  */
 115 #define WIDE_OUTPUT (1u << 28)
 116
 117   /* Use internally by the target specific disassembly code.  */
 118   void *private_data;
 119
 120   /* Function used to get bytes to disassemble.  MEMADDR is the
 121      address of the stuff to be disassembled, MYADDR is the address to
 122      put the bytes in, and LENGTH is the number of bytes to read.
 123      INFO is a pointer to this struct.
 124      Returns an errno value or 0 for success.  */
 125   int (*read_memory_func)
 126     (bfd_vma memaddr, bfd_byte *myaddr, unsigned int length,
 127      struct disassemble_info *dinfo);
 128
 129   /* Function which should be called if we get an error that we can't
 130      recover from.  STATUS is the errno value from read_memory_func and
 131      MEMADDR is the address that we were trying to read.  INFO is a
 132      pointer to this struct.  */
 133   void (*memory_error_func)
 134     (int status, bfd_vma memaddr, struct disassemble_info *dinfo);
 135
 136   /* Function called to print ADDR.  */
 137   void (*print_address_func)
 138     (bfd_vma addr, struct disassemble_info *dinfo);
 139
 140   /* Function called to determine if there is a symbol at the given ADDR.
 141      If there is, the function returns 1, otherwise it returns 0.
 142      This is used by ports which support an overlay manager where
 143      the overlay number is held in the top part of an address.  In
 144      some circumstances we want to include the overlay number in the
 145      address, (normally because there is a symbol associated with
 146      that address), but sometimes we want to mask out the overlay bits.  */
 147   int (* symbol_at_address_func)
 148     (bfd_vma addr, struct disassemble_info *dinfo);
 149
 150   /* Function called to check if a SYMBOL is can be displayed to the user.
 151      This is used by some ports that want to hide special symbols when
 152      displaying debugging outout.  */
 153   bfd_boolean (* symbol_is_valid)
 154     (asymbol *, struct disassemble_info *dinfo);
 155
 156   /* These are for buffer_read_memory.  */
 157   bfd_byte *buffer;
 158   bfd_vma buffer_vma;
 159   size_t buffer_length;
 160
 161   /* This variable may be set by the instruction decoder.  It suggests
 162       the number of bytes objdump should display on a single line.  If
 163       the instruction decoder sets this, it should always set it to
 164       the same value in order to get reasonable looking output.  */
 165   int bytes_per_line;
 166
 167   /* The next two variables control the way objdump displays the raw data.  */
 168   /* For example, if bytes_per_line is 8 and bytes_per_chunk is 4, the */
 169   /* output will look like this:
 170      00:   00000000 00000000
 171      with the chunks displayed according to "display_endian". */
 172   int bytes_per_chunk;
 173   enum bfd_endian display_endian;
 174
 175   /* Number of octets per incremented target address
 176      Normally one, but some DSPs have byte sizes of 16 or 32 bits.  */
 177   unsigned int octets_per_byte;
 178
 179   /* The number of zeroes we want to see at the end of a section before we
 180      start skipping them.  */
 181   unsigned int skip_zeroes;
 182
 183   /* The number of zeroes to skip at the end of a section.  If the number
 184      of zeroes at the end is between SKIP_ZEROES_AT_END and SKIP_ZEROES,
 185      they will be disassembled.  If there are fewer than
 186      SKIP_ZEROES_AT_END, they will be skipped.  This is a heuristic
 187      attempt to avoid disassembling zeroes inserted by section
 188      alignment.  */
 189   unsigned int skip_zeroes_at_end;
 190
 191   /* Whether the disassembler always needs the relocations.  */
 192   bfd_boolean disassembler_needs_relocs;
 193
 194   /* Results from instruction decoders.  Not all decoders yet support
 195      this information.  This info is set each time an instruction is
 196      decoded, and is only valid for the last such instruction.
 197
 198      To determine whether this decoder supports this information, set
 199      insn_info_valid to 0, decode an instruction, then check it.  */
 200
 201   char insn_info_valid;         /* Branch info has been set. */
 202   char branch_delay_insns;      /* How many sequential insn's will run before
 203                                    a branch takes effect.  (0 = normal) */
 204   char data_size;               /* Size of data reference in insn, in bytes */
 205   enum dis_insn_type insn_type; /* Type of instruction */
 206   bfd_vma target;               /* Target address of branch or dref, if known;
 207                                    zero if unknown.  */
 208   bfd_vma target2;              /* Second target address for dref2 */
 209
 210   /* Command line options specific to the target disassembler.  */
 211   const char *disassembler_options;
 212
 213   /* If non-zero then try not disassemble beyond this address, even if
 214      there are values left in the buffer.  This address is the address
 215      of the nearest symbol forwards from the start of the disassembly,
 216      and it is assumed that it lies on the boundary between instructions.
 217      If an instruction spans this address then this is an error in the
 218      file being disassembled.  */
 219   bfd_vma stop_vma;
 220
 221   /* The end range of the current range being disassembled.  This is required
 222      in order to notify the disassembler when it's currently handling a
 223      different range than it was before.  This prevent unsafe optimizations when
 224      disassembling such as the way mapping symbols are found on AArch64.  */
 225   bfd_vma stop_offset;
 226
 227 } disassemble_info;
 228
 229 /* This struct is used to pass information about valid disassembler
 230    option arguments from the target to the generic GDB functions
 231    that set and display them.  */
 232
 233 typedef struct
 234 {
 235   /* Option argument name to use in descriptions.  */
 236   const char *name;
 237
 238   /* Vector of acceptable option argument values, NULL-terminated.  */
 239   const char **values;
 240 } disasm_option_arg_t;
 241
 242 /* This struct is used to pass information about valid disassembler
 243    options, their descriptions and arguments from the target to the
 244    generic GDB functions that set and display them.  Options are
 245    defined by tuples of vector entries at each index.  */
 246
 247 typedef struct
 248 {
 249   /* Vector of option names, NULL-terminated.  */
 250   const char **name;
 251
 252   /* Vector of option descriptions or NULL if none to be shown.  */
 253   const char **description;
 254
 255   /* Vector of option argument information pointers or NULL if no
 256      option accepts an argument.  NULL entries denote individual
 257      options that accept no argument.  */
 258   const disasm_option_arg_t **arg;
 259 } disasm_options_t;
 260
 261 /* This struct is used to pass information about valid disassembler
 262    options and arguments from the target to the generic GDB functions
 263    that set and display them.  */
 264
 265 typedef struct
 266 {
 267   /* Valid disassembler options.  Individual options that support
 268      an argument will refer to entries in the ARGS vector.  */
 269   disasm_options_t options;
 270
 271   /* Vector of acceptable option arguments, NULL-terminated.  This
 272      collects all possible option argument choices, some of which
 273      may be shared by different options from the OPTIONS member.  */
 274   disasm_option_arg_t *args;
 275 } disasm_options_and_args_t;
 276 \f
 277 /* Standard disassemblers.  Disassemble one instruction at the given
 278    target address.  Return number of octets processed.  */
 279 typedef int (*disassembler_ftype) (bfd_vma, disassemble_info *);
 280
 281 /* Disassemblers used out side of opcodes library.  */
 282 extern int print_insn_m32c              (bfd_vma, disassemble_info *);
 283 extern int print_insn_mep               (bfd_vma, disassemble_info *);
 284 extern int print_insn_s12z              (bfd_vma, disassemble_info *);
 285 extern int print_insn_sh                (bfd_vma, disassemble_info *);
 286 extern int print_insn_sparc             (bfd_vma, disassemble_info *);
 287 extern int print_insn_rx                (bfd_vma, disassemble_info *);
 288 extern int print_insn_rl78              (bfd_vma, disassemble_info *);
 289 extern int print_insn_rl78_g10          (bfd_vma, disassemble_info *);
 290 extern int print_insn_rl78_g13          (bfd_vma, disassemble_info *);
 291 extern int print_insn_rl78_g14          (bfd_vma, disassemble_info *);
 292
 293 extern disassembler_ftype arc_get_disassembler (bfd *);
 294 extern disassembler_ftype cris_get_disassembler (bfd *);
 295
 296 extern void print_aarch64_disassembler_options (FILE *);
 297 extern void print_i386_disassembler_options (FILE *);
 298 extern void print_mips_disassembler_options (FILE *);
 299 extern void print_nfp_disassembler_options (FILE *);
 300 extern void print_ppc_disassembler_options (FILE *);
 301 extern void print_riscv_disassembler_options (FILE *);
 302 extern void print_arm_disassembler_options (FILE *);
 303 extern void print_arc_disassembler_options (FILE *);
 304 extern void print_s390_disassembler_options (FILE *);
 305 extern void print_wasm32_disassembler_options (FILE *);
 306 extern bfd_boolean aarch64_symbol_is_valid (asymbol *, struct disassemble_info *);
 307 extern bfd_boolean arm_symbol_is_valid (asymbol *, struct disassemble_info *);
 308 extern bfd_boolean csky_symbol_is_valid (asymbol *, struct disassemble_info *);
 309 extern bfd_boolean riscv_symbol_is_valid (asymbol *, struct disassemble_info *);
 310 extern void disassemble_init_powerpc (struct disassemble_info *);
 311 extern void disassemble_init_s390 (struct disassemble_info *);
 312 extern void disassemble_init_wasm32 (struct disassemble_info *);
 313 extern void disassemble_init_nds32 (struct disassemble_info *);
 314 extern const disasm_options_and_args_t *disassembler_options_arm (void);
 315 extern const disasm_options_and_args_t *disassembler_options_mips (void);
 316 extern const disasm_options_and_args_t *disassembler_options_powerpc (void);
 317 extern const disasm_options_and_args_t *disassembler_options_s390 (void);
 318
 319 /* Fetch the disassembler for a given architecture ARC, endianess (big
 320    endian if BIG is true), bfd_mach value MACH, and ABFD, if that support
 321    is available.  ABFD may be NULL.  */
 322 extern disassembler_ftype disassembler (enum bfd_architecture arc,
 323                                         bfd_boolean big, unsigned long mach,
 324                                         bfd *abfd);
 325
 326 /* Amend the disassemble_info structure as necessary for the target architecture.
 327    Should only be called after initialising the info->arch field.  */
 328 extern void disassemble_init_for_target (struct disassemble_info *);
 329
 330 /* Tidy any memory allocated by targets, such as info->private_data.  */
 331 extern void disassemble_free_target (struct disassemble_info *);
 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
 396 #ifdef __cplusplus
 397 }
 398 #endif
 399
 400 #endif /* ! defined (DIS_ASM_H) */