Subversion Repositories openrisc

/* This is the Assembler Pre-Processor

/* This is the Assembler Pre-Processor

   Copyright 1987, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998,

   Copyright 1987, 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998,

   1999, 2000, 2001, 2002, 2003, 2006, 2007

   1999, 2000, 2001, 2002, 2003, 2006, 2007

   Free Software Foundation, Inc.

   Free Software Foundation, Inc.

   This file is part of GAS, the GNU Assembler.

   This file is part of GAS, the GNU Assembler.

   GAS is free software; you can redistribute it and/or modify

   GAS is free software; you can redistribute it and/or modify

   it under the terms of the GNU General Public License as published by

   it under the terms of the GNU General Public License as published by

   the Free Software Foundation; either version 3, or (at your option)

   the Free Software Foundation; either version 3, or (at your option)

   any later version.

   any later version.

   GAS is distributed in the hope that it will be useful, but WITHOUT

   GAS is distributed in the hope that it will be useful, but WITHOUT

   ANY WARRANTY; without even the implied warranty of MERCHANTABILITY

   ANY WARRANTY; without even the implied warranty of MERCHANTABILITY

   or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public

   or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public

   License for more details.

   License for more details.

   You should have received a copy of the GNU General Public License

   You should have received a copy of the GNU General Public License

   along with GAS; see the file COPYING.  If not, write to the Free

   along with GAS; see the file COPYING.  If not, write to the Free

   Software Foundation, 51 Franklin Street - Fifth Floor, Boston, MA

   Software Foundation, 51 Franklin Street - Fifth Floor, Boston, MA

   02110-1301, USA.  */

   02110-1301, USA.  */

/* Modified by Allen Wirfs-Brock, Instantiations Inc 2/90.  */

/* Modified by Allen Wirfs-Brock, Instantiations Inc 2/90.  */

/* App, the assembler pre-processor.  This pre-processor strips out

/* App, the assembler pre-processor.  This pre-processor strips out

   excess spaces, turns single-quoted characters into a decimal

   excess spaces, turns single-quoted characters into a decimal

   constant, and turns the # in # <number> <filename> <garbage> into a

   constant, and turns the # in # <number> <filename> <garbage> into a

   .linefile.  This needs better error-handling.  */

   .linefile.  This needs better error-handling.  */

#include "as.h"

#include "as.h"

#if (__STDC__ != 1)

#if (__STDC__ != 1)

#ifndef const

#ifndef const

#define const  /* empty */

#define const  /* empty */

#endif

#endif

#endif

#endif

#ifdef TC_M68K

#ifdef TC_M68K

/* Whether we are scrubbing in m68k MRI mode.  This is different from

/* Whether we are scrubbing in m68k MRI mode.  This is different from

   flag_m68k_mri, because the two flags will be affected by the .mri

   flag_m68k_mri, because the two flags will be affected by the .mri

   pseudo-op at different times.  */

   pseudo-op at different times.  */

static int scrub_m68k_mri;

static int scrub_m68k_mri;

/* The pseudo-op which switches in and out of MRI mode.  See the

/* The pseudo-op which switches in and out of MRI mode.  See the

   comment in do_scrub_chars.  */

   comment in do_scrub_chars.  */

static const char mri_pseudo[] = ".mri 0";

static const char mri_pseudo[] = ".mri 0";

#else

#else

#define scrub_m68k_mri 0

#define scrub_m68k_mri 0

#endif

#endif

#if defined TC_ARM && defined OBJ_ELF

#if defined TC_ARM && defined OBJ_ELF

/* The pseudo-op for which we need to special-case `@' characters.

/* The pseudo-op for which we need to special-case `@' characters.

   See the comment in do_scrub_chars.  */

   See the comment in do_scrub_chars.  */

static const char   symver_pseudo[] = ".symver";

static const char   symver_pseudo[] = ".symver";

static const char * symver_state;

static const char * symver_state;

#endif

#endif

static char lex[256];

static char lex[256];

static const char symbol_chars[] =

static const char symbol_chars[] =

"$._ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789";

"$._ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789";

#define LEX_IS_SYMBOL_COMPONENT         1

#define LEX_IS_SYMBOL_COMPONENT         1

#define LEX_IS_WHITESPACE               2

#define LEX_IS_WHITESPACE               2

#define LEX_IS_LINE_SEPARATOR           3

#define LEX_IS_LINE_SEPARATOR           3

#define LEX_IS_COMMENT_START            4

#define LEX_IS_COMMENT_START            4

#define LEX_IS_LINE_COMMENT_START       5

#define LEX_IS_LINE_COMMENT_START       5

#define LEX_IS_TWOCHAR_COMMENT_1ST      6

#define LEX_IS_TWOCHAR_COMMENT_1ST      6

#define LEX_IS_STRINGQUOTE              8

#define LEX_IS_STRINGQUOTE              8

#define LEX_IS_COLON                    9

#define LEX_IS_COLON                    9

#define LEX_IS_NEWLINE                  10

#define LEX_IS_NEWLINE                  10

#define LEX_IS_ONECHAR_QUOTE            11

#define LEX_IS_ONECHAR_QUOTE            11

#ifdef TC_V850

#ifdef TC_V850

#define LEX_IS_DOUBLEDASH_1ST           12

#define LEX_IS_DOUBLEDASH_1ST           12

#endif

#endif

#ifdef TC_M32R

#ifdef TC_M32R

#define DOUBLEBAR_PARALLEL

#define DOUBLEBAR_PARALLEL

#endif

#endif

#ifdef DOUBLEBAR_PARALLEL

#ifdef DOUBLEBAR_PARALLEL

#define LEX_IS_DOUBLEBAR_1ST            13

#define LEX_IS_DOUBLEBAR_1ST            13

#endif

#endif

#define LEX_IS_PARALLEL_SEPARATOR       14

#define LEX_IS_PARALLEL_SEPARATOR       14

#define IS_SYMBOL_COMPONENT(c)          (lex[c] == LEX_IS_SYMBOL_COMPONENT)

#define IS_SYMBOL_COMPONENT(c)          (lex[c] == LEX_IS_SYMBOL_COMPONENT)

#define IS_WHITESPACE(c)                (lex[c] == LEX_IS_WHITESPACE)

#define IS_WHITESPACE(c)                (lex[c] == LEX_IS_WHITESPACE)

#define IS_LINE_SEPARATOR(c)            (lex[c] == LEX_IS_LINE_SEPARATOR)

#define IS_LINE_SEPARATOR(c)            (lex[c] == LEX_IS_LINE_SEPARATOR)

#define IS_PARALLEL_SEPARATOR(c)        (lex[c] == LEX_IS_PARALLEL_SEPARATOR)

#define IS_PARALLEL_SEPARATOR(c)        (lex[c] == LEX_IS_PARALLEL_SEPARATOR)

#define IS_COMMENT(c)                   (lex[c] == LEX_IS_COMMENT_START)

#define IS_COMMENT(c)                   (lex[c] == LEX_IS_COMMENT_START)

#define IS_LINE_COMMENT(c)              (lex[c] == LEX_IS_LINE_COMMENT_START)

#define IS_LINE_COMMENT(c)              (lex[c] == LEX_IS_LINE_COMMENT_START)

#define IS_NEWLINE(c)                   (lex[c] == LEX_IS_NEWLINE)

#define IS_NEWLINE(c)                   (lex[c] == LEX_IS_NEWLINE)

static int process_escape (int);

static int process_escape (int);

/* FIXME-soon: The entire lexer/parser thingy should be

/* FIXME-soon: The entire lexer/parser thingy should be

   built statically at compile time rather than dynamically

   built statically at compile time rather than dynamically

   each and every time the assembler is run.  xoxorich.  */

   each and every time the assembler is run.  xoxorich.  */

void

void

do_scrub_begin (int m68k_mri ATTRIBUTE_UNUSED)

do_scrub_begin (int m68k_mri ATTRIBUTE_UNUSED)

{

{

  const char *p;

  const char *p;

  int c;

  int c;

  lex[' '] = LEX_IS_WHITESPACE;

  lex[' '] = LEX_IS_WHITESPACE;

  lex['\t'] = LEX_IS_WHITESPACE;

  lex['\t'] = LEX_IS_WHITESPACE;

  lex['\r'] = LEX_IS_WHITESPACE;

  lex['\r'] = LEX_IS_WHITESPACE;

  lex['\n'] = LEX_IS_NEWLINE;

  lex['\n'] = LEX_IS_NEWLINE;

  lex[':'] = LEX_IS_COLON;

  lex[':'] = LEX_IS_COLON;

#ifdef TC_M68K

#ifdef TC_M68K

  scrub_m68k_mri = m68k_mri;

  scrub_m68k_mri = m68k_mri;

  if (! m68k_mri)

  if (! m68k_mri)

#endif

#endif

    {

    {

      lex['"'] = LEX_IS_STRINGQUOTE;

      lex['"'] = LEX_IS_STRINGQUOTE;

#if ! defined (TC_HPPA) && ! defined (TC_I370)

#if ! defined (TC_HPPA) && ! defined (TC_I370)

      /* I370 uses single-quotes to delimit integer, float constants.  */

      /* I370 uses single-quotes to delimit integer, float constants.  */

      lex['\''] = LEX_IS_ONECHAR_QUOTE;

      lex['\''] = LEX_IS_ONECHAR_QUOTE;

#endif

#endif

#ifdef SINGLE_QUOTE_STRINGS

#ifdef SINGLE_QUOTE_STRINGS

      lex['\''] = LEX_IS_STRINGQUOTE;

      lex['\''] = LEX_IS_STRINGQUOTE;

#endif

#endif

    }

    }

  /* Note: if any other character can be LEX_IS_STRINGQUOTE, the loop

  /* Note: if any other character can be LEX_IS_STRINGQUOTE, the loop

     in state 5 of do_scrub_chars must be changed.  */

     in state 5 of do_scrub_chars must be changed.  */

  /* Note that these override the previous defaults, e.g. if ';' is a

  /* Note that these override the previous defaults, e.g. if ';' is a

     comment char, then it isn't a line separator.  */

     comment char, then it isn't a line separator.  */

  for (p = symbol_chars; *p; ++p)

  for (p = symbol_chars; *p; ++p)

    lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT;

    lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT;

  for (c = 128; c < 256; ++c)

  for (c = 128; c < 256; ++c)

    lex[c] = LEX_IS_SYMBOL_COMPONENT;

    lex[c] = LEX_IS_SYMBOL_COMPONENT;

#ifdef tc_symbol_chars

#ifdef tc_symbol_chars

  /* This macro permits the processor to specify all characters which

  /* This macro permits the processor to specify all characters which

     may appears in an operand.  This will prevent the scrubber from

     may appears in an operand.  This will prevent the scrubber from

     discarding meaningful whitespace in certain cases.  The i386

     discarding meaningful whitespace in certain cases.  The i386

     backend uses this to support prefixes, which can confuse the

     backend uses this to support prefixes, which can confuse the

     scrubber as to whether it is parsing operands or opcodes.  */

     scrubber as to whether it is parsing operands or opcodes.  */

  for (p = tc_symbol_chars; *p; ++p)

  for (p = tc_symbol_chars; *p; ++p)

    lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT;

    lex[(unsigned char) *p] = LEX_IS_SYMBOL_COMPONENT;

#endif

#endif

  /* The m68k backend wants to be able to change comment_chars.  */

  /* The m68k backend wants to be able to change comment_chars.  */

#ifndef tc_comment_chars

#ifndef tc_comment_chars

#define tc_comment_chars comment_chars

#define tc_comment_chars comment_chars

#endif

#endif

  for (p = tc_comment_chars; *p; p++)

  for (p = tc_comment_chars; *p; p++)

    lex[(unsigned char) *p] = LEX_IS_COMMENT_START;

    lex[(unsigned char) *p] = LEX_IS_COMMENT_START;

  for (p = line_comment_chars; *p; p++)

  for (p = line_comment_chars; *p; p++)

    lex[(unsigned char) *p] = LEX_IS_LINE_COMMENT_START;

    lex[(unsigned char) *p] = LEX_IS_LINE_COMMENT_START;

  for (p = line_separator_chars; *p; p++)

  for (p = line_separator_chars; *p; p++)

    lex[(unsigned char) *p] = LEX_IS_LINE_SEPARATOR;

    lex[(unsigned char) *p] = LEX_IS_LINE_SEPARATOR;

#ifdef tc_parallel_separator_chars

#ifdef tc_parallel_separator_chars

  /* This macro permits the processor to specify all characters which

  /* This macro permits the processor to specify all characters which

     separate parallel insns on the same line.  */

     separate parallel insns on the same line.  */

  for (p = tc_parallel_separator_chars; *p; p++)

  for (p = tc_parallel_separator_chars; *p; p++)

    lex[(unsigned char) *p] = LEX_IS_PARALLEL_SEPARATOR;

    lex[(unsigned char) *p] = LEX_IS_PARALLEL_SEPARATOR;

#endif

#endif

  /* Only allow slash-star comments if slash is not in use.

  /* Only allow slash-star comments if slash is not in use.

     FIXME: This isn't right.  We should always permit them.  */

     FIXME: This isn't right.  We should always permit them.  */

  if (lex['/'] == 0)

  if (lex['/'] == 0)

    lex['/'] = LEX_IS_TWOCHAR_COMMENT_1ST;

    lex['/'] = LEX_IS_TWOCHAR_COMMENT_1ST;

#ifdef TC_M68K

#ifdef TC_M68K

  if (m68k_mri)

  if (m68k_mri)

    {

    {

      lex['\''] = LEX_IS_STRINGQUOTE;

      lex['\''] = LEX_IS_STRINGQUOTE;

      lex[';'] = LEX_IS_COMMENT_START;

      lex[';'] = LEX_IS_COMMENT_START;

      lex['*'] = LEX_IS_LINE_COMMENT_START;

      lex['*'] = LEX_IS_LINE_COMMENT_START;

      /* The MRI documentation says '!' is LEX_IS_COMMENT_START, but

      /* The MRI documentation says '!' is LEX_IS_COMMENT_START, but

         then it can't be used in an expression.  */

         then it can't be used in an expression.  */

      lex['!'] = LEX_IS_LINE_COMMENT_START;

      lex['!'] = LEX_IS_LINE_COMMENT_START;

    }

    }

#endif

#endif

#ifdef TC_V850

#ifdef TC_V850

  lex['-'] = LEX_IS_DOUBLEDASH_1ST;

  lex['-'] = LEX_IS_DOUBLEDASH_1ST;

#endif

#endif

#ifdef DOUBLEBAR_PARALLEL

#ifdef DOUBLEBAR_PARALLEL

  lex['|'] = LEX_IS_DOUBLEBAR_1ST;

  lex['|'] = LEX_IS_DOUBLEBAR_1ST;

#endif

#endif

#ifdef TC_D30V

#ifdef TC_D30V

  /* Must do this is we want VLIW instruction with "->" or "<-".  */

  /* Must do this is we want VLIW instruction with "->" or "<-".  */

  lex['-'] = LEX_IS_SYMBOL_COMPONENT;

  lex['-'] = LEX_IS_SYMBOL_COMPONENT;

#endif

#endif

}

}

/* Saved state of the scrubber.  */

/* Saved state of the scrubber.  */

static int state;

static int state;

static int old_state;

static int old_state;

static char *out_string;

static char *out_string;

static char out_buf[20];

static char out_buf[20];

static int add_newlines;

static int add_newlines;

static char *saved_input;

static char *saved_input;

static int saved_input_len;

static int saved_input_len;

static char input_buffer[32 * 1024];

static char input_buffer[32 * 1024];

static const char *mri_state;

static const char *mri_state;

static char mri_last_ch;

static char mri_last_ch;

/* Data structure for saving the state of app across #include's.  Note that

/* Data structure for saving the state of app across #include's.  Note that

   app is called asynchronously to the parsing of the .include's, so our

   app is called asynchronously to the parsing of the .include's, so our

   state at the time .include is interpreted is completely unrelated.

   state at the time .include is interpreted is completely unrelated.

   That's why we have to save it all.  */

   That's why we have to save it all.  */

struct app_save

struct app_save

{

{

  int          state;

  int          state;

  int          old_state;

  int          old_state;

  char *       out_string;

  char *       out_string;

  char         out_buf[sizeof (out_buf)];

  char         out_buf[sizeof (out_buf)];

  int          add_newlines;

  int          add_newlines;

  char *       saved_input;

  char *       saved_input;

  int          saved_input_len;

  int          saved_input_len;

#ifdef TC_M68K

#ifdef TC_M68K

  int          scrub_m68k_mri;

  int          scrub_m68k_mri;

#endif

#endif

  const char * mri_state;

  const char * mri_state;

  char         mri_last_ch;

  char         mri_last_ch;

#if defined TC_ARM && defined OBJ_ELF

#if defined TC_ARM && defined OBJ_ELF

  const char * symver_state;

  const char * symver_state;

#endif

#endif

};

};

char *

char *

app_push (void)

app_push (void)

{

{

  register struct app_save *saved;

  register struct app_save *saved;

  saved = (struct app_save *) xmalloc (sizeof (*saved));

  saved = (struct app_save *) xmalloc (sizeof (*saved));

  saved->state = state;

  saved->state = state;

  saved->old_state = old_state;

  saved->old_state = old_state;

  saved->out_string = out_string;

  saved->out_string = out_string;

  memcpy (saved->out_buf, out_buf, sizeof (out_buf));

  memcpy (saved->out_buf, out_buf, sizeof (out_buf));

  saved->add_newlines = add_newlines;

  saved->add_newlines = add_newlines;

  if (saved_input == NULL)

  if (saved_input == NULL)

    saved->saved_input = NULL;

    saved->saved_input = NULL;

  else

  else

    {

    {

      saved->saved_input = xmalloc (saved_input_len);

      saved->saved_input = xmalloc (saved_input_len);

      memcpy (saved->saved_input, saved_input, saved_input_len);

      memcpy (saved->saved_input, saved_input, saved_input_len);

      saved->saved_input_len = saved_input_len;

      saved->saved_input_len = saved_input_len;

    }

    }

#ifdef TC_M68K

#ifdef TC_M68K

  saved->scrub_m68k_mri = scrub_m68k_mri;

  saved->scrub_m68k_mri = scrub_m68k_mri;

#endif

#endif

  saved->mri_state = mri_state;

  saved->mri_state = mri_state;

  saved->mri_last_ch = mri_last_ch;

  saved->mri_last_ch = mri_last_ch;

#if defined TC_ARM && defined OBJ_ELF

#if defined TC_ARM && defined OBJ_ELF

  saved->symver_state = symver_state;

  saved->symver_state = symver_state;

#endif

#endif

  /* do_scrub_begin() is not useful, just wastes time.  */

  /* do_scrub_begin() is not useful, just wastes time.  */

  state = 0;

  state = 0;

  saved_input = NULL;

  saved_input = NULL;

  return (char *) saved;

  return (char *) saved;

}

}

void

void

app_pop (char *arg)

app_pop (char *arg)

{

{

  register struct app_save *saved = (struct app_save *) arg;

  register struct app_save *saved = (struct app_save *) arg;

  /* There is no do_scrub_end ().  */

  /* There is no do_scrub_end ().  */

  state = saved->state;

  state = saved->state;

  old_state = saved->old_state;

  old_state = saved->old_state;

  out_string = saved->out_string;

  out_string = saved->out_string;

  memcpy (out_buf, saved->out_buf, sizeof (out_buf));

  memcpy (out_buf, saved->out_buf, sizeof (out_buf));

  add_newlines = saved->add_newlines;

  add_newlines = saved->add_newlines;

  if (saved->saved_input == NULL)

  if (saved->saved_input == NULL)

    saved_input = NULL;

    saved_input = NULL;

  else

  else

    {

    {

      assert (saved->saved_input_len <= (int) (sizeof input_buffer));

      assert (saved->saved_input_len <= (int) (sizeof input_buffer));

      memcpy (input_buffer, saved->saved_input, saved->saved_input_len);

      memcpy (input_buffer, saved->saved_input, saved->saved_input_len);

      saved_input = input_buffer;

      saved_input = input_buffer;

      saved_input_len = saved->saved_input_len;

      saved_input_len = saved->saved_input_len;

      free (saved->saved_input);

      free (saved->saved_input);

    }

    }

#ifdef TC_M68K

#ifdef TC_M68K

  scrub_m68k_mri = saved->scrub_m68k_mri;

  scrub_m68k_mri = saved->scrub_m68k_mri;

#endif

#endif

  mri_state = saved->mri_state;

  mri_state = saved->mri_state;

  mri_last_ch = saved->mri_last_ch;

  mri_last_ch = saved->mri_last_ch;

#if defined TC_ARM && defined OBJ_ELF

#if defined TC_ARM && defined OBJ_ELF

  symver_state = saved->symver_state;

  symver_state = saved->symver_state;

#endif

#endif

  free (arg);

  free (arg);

}

}

/* @@ This assumes that \n &c are the same on host and target.  This is not

/* @@ This assumes that \n &c are the same on host and target.  This is not

   necessarily true.  */

   necessarily true.  */

static int

static int

process_escape (int ch)

process_escape (int ch)

{

{

  switch (ch)

  switch (ch)

    {

    {

    case 'b':

    case 'b':

      return '\b';

      return '\b';

    case 'f':

    case 'f':

      return '\f';

      return '\f';

    case 'n':

    case 'n':

      return '\n';

      return '\n';

    case 'r':

    case 'r':

      return '\r';

      return '\r';

    case 't':

    case 't':

      return '\t';

      return '\t';

    case '\'':

    case '\'':

      return '\'';

      return '\'';

    case '"':

    case '"':

      return '\"';

      return '\"';

    default:

    default:

      return ch;

      return ch;

    }

    }

}

}

/* This function is called to process input characters.  The GET

/* This function is called to process input characters.  The GET

   parameter is used to retrieve more input characters.  GET should

   parameter is used to retrieve more input characters.  GET should

   set its parameter to point to a buffer, and return the length of

   set its parameter to point to a buffer, and return the length of

   the buffer; it should return 0 at end of file.  The scrubbed output

   the buffer; it should return 0 at end of file.  The scrubbed output

   characters are put into the buffer starting at TOSTART; the TOSTART

   characters are put into the buffer starting at TOSTART; the TOSTART

   buffer is TOLEN bytes in length.  The function returns the number

   buffer is TOLEN bytes in length.  The function returns the number

   of scrubbed characters put into TOSTART.  This will be TOLEN unless

   of scrubbed characters put into TOSTART.  This will be TOLEN unless

   end of file was seen.  This function is arranged as a state

   end of file was seen.  This function is arranged as a state

   machine, and saves its state so that it may return at any point.

   machine, and saves its state so that it may return at any point.

   This is the way the old code used to work.  */

   This is the way the old code used to work.  */

int

int

do_scrub_chars (int (*get) (char *, int), char *tostart, int tolen)

do_scrub_chars (int (*get) (char *, int), char *tostart, int tolen)

{

{

  char *to = tostart;

  char *to = tostart;

  char *toend = tostart + tolen;

  char *toend = tostart + tolen;

  char *from;

  char *from;

  char *fromend;

  char *fromend;

  int fromlen;

  int fromlen;

  register int ch, ch2 = 0;

  register int ch, ch2 = 0;

  /* Character that started the string we're working on.  */

  /* Character that started the string we're working on.  */

  static char quotechar;

  static char quotechar;

  /*State 0: beginning of normal line

  /*State 0: beginning of normal line

          1: After first whitespace on line (flush more white)

          1: After first whitespace on line (flush more white)

          2: After first non-white (opcode) on line (keep 1white)

          2: After first non-white (opcode) on line (keep 1white)

          3: after second white on line (into operands) (flush white)

          3: after second white on line (into operands) (flush white)

          4: after putting out a .linefile, put out digits

          4: after putting out a .linefile, put out digits

          5: parsing a string, then go to old-state

          5: parsing a string, then go to old-state

          6: putting out \ escape in a "d string.

          6: putting out \ escape in a "d string.

          7: no longer used

          7: no longer used

          8: no longer used

          8: no longer used

          9: After seeing symbol char in state 3 (keep 1white after symchar)

          9: After seeing symbol char in state 3 (keep 1white after symchar)

         10: After seeing whitespace in state 9 (keep white before symchar)

         10: After seeing whitespace in state 9 (keep white before symchar)

         11: After seeing a symbol character in state 0 (eg a label definition)

         11: After seeing a symbol character in state 0 (eg a label definition)

         -1: output string in out_string and go to the state in old_state

         -1: output string in out_string and go to the state in old_state

         -2: flush text until a '*' '/' is seen, then go to state old_state

         -2: flush text until a '*' '/' is seen, then go to state old_state

#ifdef TC_V850

#ifdef TC_V850

         12: After seeing a dash, looking for a second dash as a start

         12: After seeing a dash, looking for a second dash as a start

             of comment.

             of comment.

#endif

#endif

#ifdef DOUBLEBAR_PARALLEL

#ifdef DOUBLEBAR_PARALLEL

         13: After seeing a vertical bar, looking for a second

         13: After seeing a vertical bar, looking for a second

             vertical bar as a parallel expression separator.

             vertical bar as a parallel expression separator.

#endif

#endif

#ifdef TC_IA64

#ifdef TC_IA64

         14: After seeing a `(' at state 0, looking for a `)' as

         14: After seeing a `(' at state 0, looking for a `)' as

             predicate.

             predicate.

         15: After seeing a `(' at state 1, looking for a `)' as

         15: After seeing a `(' at state 1, looking for a `)' as

             predicate.

             predicate.

#endif

#endif

#ifdef TC_Z80

#ifdef TC_Z80

         16: After seeing an 'a' or an 'A' at the start of a symbol

         16: After seeing an 'a' or an 'A' at the start of a symbol

         17: After seeing an 'f' or an 'F' in state 16

         17: After seeing an 'f' or an 'F' in state 16

#endif

#endif

          */

          */

  /* I added states 9 and 10 because the MIPS ECOFF assembler uses

  /* I added states 9 and 10 because the MIPS ECOFF assembler uses

     constructs like ``.loc 1 20''.  This was turning into ``.loc

     constructs like ``.loc 1 20''.  This was turning into ``.loc

     120''.  States 9 and 10 ensure that a space is never dropped in

     120''.  States 9 and 10 ensure that a space is never dropped in

     between characters which could appear in an identifier.  Ian

     between characters which could appear in an identifier.  Ian

     Taylor, ian@cygnus.com.

     Taylor, ian@cygnus.com.

     I added state 11 so that something like "Lfoo add %r25,%r26,%r27" works

     I added state 11 so that something like "Lfoo add %r25,%r26,%r27" works

     correctly on the PA (and any other target where colons are optional).

     correctly on the PA (and any other target where colons are optional).

     Jeff Law, law@cs.utah.edu.

     Jeff Law, law@cs.utah.edu.

     I added state 13 so that something like "cmp r1, r2 || trap #1" does not

     I added state 13 so that something like "cmp r1, r2 || trap #1" does not

     get squashed into "cmp r1,r2||trap#1", with the all important space

     get squashed into "cmp r1,r2||trap#1", with the all important space

     between the 'trap' and the '#1' being eliminated.  nickc@cygnus.com  */

     between the 'trap' and the '#1' being eliminated.  nickc@cygnus.com  */

  /* This macro gets the next input character.  */

  /* This macro gets the next input character.  */

#define GET()                                                   \

#define GET()                                                   \

  (from < fromend                                               \

  (from < fromend                                               \

   ? * (unsigned char *) (from++)                               \

   ? * (unsigned char *) (from++)                               \

   : (saved_input = NULL,                                       \

   : (saved_input = NULL,                                       \

      fromlen = (*get) (input_buffer, sizeof input_buffer),     \

      fromlen = (*get) (input_buffer, sizeof input_buffer),     \

      from = input_buffer,                                      \

      from = input_buffer,                                      \

      fromend = from + fromlen,                                 \

      fromend = from + fromlen,                                 \

      (fromlen == 0                                              \

      (fromlen == 0                                              \

       ? EOF                                                    \

       ? EOF                                                    \

       : * (unsigned char *) (from++))))

       : * (unsigned char *) (from++))))

  /* This macro pushes a character back on the input stream.  */

  /* This macro pushes a character back on the input stream.  */

#define UNGET(uch) (*--from = (uch))

#define UNGET(uch) (*--from = (uch))

  /* This macro puts a character into the output buffer.  If this

  /* This macro puts a character into the output buffer.  If this

     character fills the output buffer, this macro jumps to the label

     character fills the output buffer, this macro jumps to the label

     TOFULL.  We use this rather ugly approach because we need to

     TOFULL.  We use this rather ugly approach because we need to

     handle two different termination conditions: EOF on the input

     handle two different termination conditions: EOF on the input

     stream, and a full output buffer.  It would be simpler if we

     stream, and a full output buffer.  It would be simpler if we

     always read in the entire input stream before processing it, but

     always read in the entire input stream before processing it, but

     I don't want to make such a significant change to the assembler's

     I don't want to make such a significant change to the assembler's

     memory usage.  */

     memory usage.  */

#define PUT(pch)                                \

#define PUT(pch)                                \

  do                                            \

  do                                            \

    {                                           \

    {                                           \

      *to++ = (pch);                            \

      *to++ = (pch);                            \

      if (to >= toend)                          \

      if (to >= toend)                          \

        goto tofull;                            \

        goto tofull;                            \

    }                                           \

    }                                           \

  while (0)

  while (0)

  if (saved_input != NULL)

  if (saved_input != NULL)

    {

    {

      from = saved_input;

      from = saved_input;

      fromend = from + saved_input_len;

      fromend = from + saved_input_len;

    }

    }

  else

  else

    {

    {

      fromlen = (*get) (input_buffer, sizeof input_buffer);

      fromlen = (*get) (input_buffer, sizeof input_buffer);

      if (fromlen == 0)

      if (fromlen == 0)

        return 0;

        return 0;

      from = input_buffer;

      from = input_buffer;

      fromend = from + fromlen;

      fromend = from + fromlen;

    }

    }

  while (1)

  while (1)

    {

    {

      /* The cases in this switch end with continue, in order to

      /* The cases in this switch end with continue, in order to

         branch back to the top of this while loop and generate the

         branch back to the top of this while loop and generate the

         next output character in the appropriate state.  */

         next output character in the appropriate state.  */

      switch (state)

      switch (state)

        {

        {

        case -1:

        case -1:

          ch = *out_string++;

          ch = *out_string++;

          if (*out_string == '\0')

          if (*out_string == '\0')

            {

            {

              state = old_state;

              state = old_state;

              old_state = 3;

              old_state = 3;

            }

            }

          PUT (ch);

          PUT (ch);

          continue;

          continue;

        case -2:

        case -2:

          for (;;)

          for (;;)

            {

            {

              do

              do

                {

                {

                  ch = GET ();

                  ch = GET ();

                  if (ch == EOF)

                  if (ch == EOF)

                    {

                    {

                      as_warn (_("end of file in comment"));

                      as_warn (_("end of file in comment"));

                      goto fromeof;

                      goto fromeof;

                    }

                    }

                  if (ch == '\n')

                  if (ch == '\n')

                    PUT ('\n');

                    PUT ('\n');

                }

                }

              while (ch != '*');

              while (ch != '*');

              while ((ch = GET ()) == '*')

              while ((ch = GET ()) == '*')

                ;

                ;

              if (ch == EOF)

              if (ch == EOF)

                {

                {

                  as_warn (_("end of file in comment"));

                  as_warn (_("end of file in comment"));

                  goto fromeof;

                  goto fromeof;

                }

                }

              if (ch == '/')

              if (ch == '/')

                break;

                break;

              UNGET (ch);

              UNGET (ch);

            }

            }

          state = old_state;

          state = old_state;

          UNGET (' ');

          UNGET (' ');

          continue;

          continue;

        case 4:

        case 4:

          ch = GET ();

          ch = GET ();

          if (ch == EOF)

          if (ch == EOF)

            goto fromeof;

            goto fromeof;

          else if (ch >= '0' && ch <= '9')

          else if (ch >= '0' && ch <= '9')

            PUT (ch);

            PUT (ch);

          else

          else

            {

            {

              while (ch != EOF && IS_WHITESPACE (ch))

              while (ch != EOF && IS_WHITESPACE (ch))

                ch = GET ();

                ch = GET ();

              if (ch == '"')

              if (ch == '"')

                {

                {

                  quotechar = ch;

                  quotechar = ch;

                  state = 5;

                  state = 5;

                  old_state = 3;

                  old_state = 3;

                  PUT (ch);

                  PUT (ch);

                }

                }

              else

              else

                {

                {

                  while (ch != EOF && ch != '\n')

                  while (ch != EOF && ch != '\n')

                    ch = GET ();

                    ch = GET ();

                  state = 0;

                  state = 0;

                  PUT (ch);

                  PUT (ch);

                }

                }

            }

            }

          continue;

          continue;

        case 5:

        case 5:

          /* We are going to copy everything up to a quote character,

          /* We are going to copy everything up to a quote character,

             with special handling for a backslash.  We try to

             with special handling for a backslash.  We try to

             optimize the copying in the simple case without using the

             optimize the copying in the simple case without using the

             GET and PUT macros.  */

             GET and PUT macros.  */

          {

          {

            char *s;

            char *s;

            int len;

            int len;

            for (s = from; s < fromend; s++)

            for (s = from; s < fromend; s++)

              {

              {

                ch = *s;

                ch = *s;

                if (ch == '\\'

                if (ch == '\\'

                    || ch == quotechar

                    || ch == quotechar

                    || ch == '\n')

                    || ch == '\n')

                  break;

                  break;

              }

              }

            len = s - from;

            len = s - from;

            if (len > toend - to)

            if (len > toend - to)

              len = toend - to;

              len = toend - to;

            if (len > 0)

            if (len > 0)

              {

              {

                memcpy (to, from, len);

                memcpy (to, from, len);

                to += len;

                to += len;

                from += len;

                from += len;

                if (to >= toend)

                if (to >= toend)

                  goto tofull;

                  goto tofull;

              }

              }

          }

          }

          ch = GET ();

          ch = GET ();

          if (ch == EOF)

          if (ch == EOF)

            {

            {

              /* This buffer is here specifically so

              /* This buffer is here specifically so

                 that the UNGET below will work.  */

                 that the UNGET below will work.  */

              static char one_char_buf[1];

              static char one_char_buf[1];

              as_warn (_("end of file in string; '%c' inserted"), quotechar);

              as_warn (_("end of file in string; '%c' inserted"), quotechar);

              state = old_state;

              state = old_state;

              from = fromend = one_char_buf + 1;

              from = fromend = one_char_buf + 1;

              fromlen = 1;

              fromlen = 1;

              UNGET ('\n');

              UNGET ('\n');

              PUT (quotechar);

              PUT (quotechar);

            }

            }

          else if (ch == quotechar)

          else if (ch == quotechar)

            {

            {

              state = old_state;

              state = old_state;

              PUT (ch);

              PUT (ch);

            }

            }

#ifndef NO_STRING_ESCAPES

#ifndef NO_STRING_ESCAPES

          else if (ch == '\\')

          else if (ch == '\\')

            {

            {

              state = 6;

              state = 6;

              PUT (ch);

              PUT (ch);

            }

            }

#endif

#endif

          else if (scrub_m68k_mri && ch == '\n')

          else if (scrub_m68k_mri && ch == '\n')

            {

            {

              /* Just quietly terminate the string.  This permits lines like

              /* Just quietly terminate the string.  This permits lines like

                   bne  label   loop if we haven't reach end yet.  */

                   bne  label   loop if we haven't reach end yet.  */

              state = old_state;

              state = old_state;

              UNGET (ch);

              UNGET (ch);

              PUT ('\'');