Change inferior thread list to be a thread map
[deliverable/binutils-gdb.git] / gdb / c-lang.c
index ad8fa5c8ebd475ecdd88f584c8b64da987a4b173..1f40e885d99aac3e41227fd22e15f7ccb40d8ce2 100644 (file)
@@ -1,12 +1,12 @@
 /* C language support routines for GDB, the GNU debugger.
 /* C language support routines for GDB, the GNU debugger.
-   Copyright 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000, 2002, 2003, 2004
-   Free Software Foundation, Inc.
+
+   Copyright (C) 1992-2019 Free Software Foundation, Inc.
 
    This file is part of GDB.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
 
    This file is part of GDB.
 
    This program is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
-   the Free Software Foundation; either version 2 of the License, or
+   the Free Software Foundation; either version 3 of the License, or
    (at your option) any later version.
 
    This program is distributed in the hope that it will be useful,
    (at your option) any later version.
 
    This program is distributed in the hope that it will be useful,
@@ -15,9 +15,7 @@
    GNU General Public License for more details.
 
    You should have received a copy of the GNU General Public License
    GNU General Public License for more details.
 
    You should have received a copy of the GNU General Public License
-   along with this program; if not, write to the Free Software
-   Foundation, Inc., 59 Temple Place - Suite 330,
-   Boston, MA 02111-1307, USA.  */
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
 
 #include "defs.h"
 #include "symtab.h"
 
 #include "defs.h"
 #include "symtab.h"
 #include "expression.h"
 #include "parser-defs.h"
 #include "language.h"
 #include "expression.h"
 #include "parser-defs.h"
 #include "language.h"
+#include "varobj.h"
 #include "c-lang.h"
 #include "c-lang.h"
+#include "c-support.h"
 #include "valprint.h"
 #include "macroscope.h"
 #include "valprint.h"
 #include "macroscope.h"
-#include "gdb_assert.h"
 #include "charset.h"
 #include "charset.h"
-#include "gdb_string.h"
 #include "demangle.h"
 #include "demangle.h"
+#include "cp-abi.h"
 #include "cp-support.h"
 #include "cp-support.h"
+#include "gdb_obstack.h"
+#include <ctype.h>
+#include "gdbcore.h"
+#include "gdbarch.h"
 
 
-extern void _initialize_c_language (void);
-static void c_emit_char (int c, struct ui_file * stream, int quoter);
-
-/* Print the character C on STREAM as part of the contents of a literal
-   string whose delimiter is QUOTER.  Note that that format for printing
-   characters and strings is language specific. */
+/* Given a C string type, STR_TYPE, return the corresponding target
+   character set name.  */
 
 
-static void
-c_emit_char (int c, struct ui_file *stream, int quoter)
+static const char *
+charset_for_string_type (c_string_type str_type, struct gdbarch *gdbarch)
 {
 {
-  const char *escape;
-  int host_char;
-
-  c &= 0xFF;                   /* Avoid sign bit follies */
-
-  escape = c_target_char_has_backslash_escape (c);
-  if (escape)
+  switch (str_type & ~C_CHAR)
     {
     {
-      if (quoter == '"' && strcmp (escape, "0") == 0)
-       /* Print nulls embedded in double quoted strings as \000 to
-          prevent ambiguity.  */
-       fprintf_filtered (stream, "\\000");
+    case C_STRING:
+      return target_charset (gdbarch);
+    case C_WIDE_STRING:
+      return target_wide_charset (gdbarch);
+    case C_STRING_16:
+      /* FIXME: UTF-16 is not always correct.  */
+      if (gdbarch_byte_order (gdbarch) == BFD_ENDIAN_BIG)
+       return "UTF-16BE";
       else
       else
-       fprintf_filtered (stream, "\\%s", escape);
-    }
-  else if (target_char_to_host (c, &host_char)
-           && host_char_print_literally (host_char))
-    {
-      if (host_char == '\\' || host_char == quoter)
-        fputs_filtered ("\\", stream);
-      fprintf_filtered (stream, "%c", host_char);
+       return "UTF-16LE";
+    case C_STRING_32:
+      /* FIXME: UTF-32 is not always correct.  */
+      if (gdbarch_byte_order (gdbarch) == BFD_ENDIAN_BIG)
+       return "UTF-32BE";
+      else
+       return "UTF-32LE";
     }
     }
-  else
-    fprintf_filtered (stream, "\\%.3o", (unsigned int) c);
-}
-
-void
-c_printchar (int c, struct ui_file *stream)
-{
-  fputc_filtered ('\'', stream);
-  LA_EMIT_CHAR (c, stream, '\'');
-  fputc_filtered ('\'', stream);
+  internal_error (__FILE__, __LINE__, _("unhandled c_string_type"));
 }
 
 }
 
-/* Print the character string STRING, printing at most LENGTH characters.
-   LENGTH is -1 if the string is nul terminated.  Each character is WIDTH bytes
-   long.  Printing stops early if the number hits print_max; repeat counts are
-   printed as appropriate.  Print ellipses at the end if we had to stop before
-   printing LENGTH characters, or if FORCE_ELLIPSES.  */
+/* Classify ELTTYPE according to what kind of character it is.  Return
+   the enum constant representing the character type.  Also set
+   *ENCODING to the name of the character set to use when converting
+   characters of this type in target BYTE_ORDER to the host character
+   set.  */
 
 
-void
-c_printstr (struct ui_file *stream, char *string, unsigned int length,
-           int width, int force_ellipses)
+static c_string_type
+classify_type (struct type *elttype, struct gdbarch *gdbarch,
+              const char **encoding)
 {
 {
-  unsigned int i;
-  unsigned int things_printed = 0;
-  int in_quotes = 0;
-  int need_comma = 0;
-
-  /* If the string was not truncated due to `set print elements', and
-     the last byte of it is a null, we don't print that, in traditional C
-     style.  */
-  if (!force_ellipses
-      && length > 0
-      && (extract_unsigned_integer (string + (length - 1) * width, width)
-          == '\0'))
-    length--;
-
-  if (length == 0)
-    {
-      fputs_filtered ("\"\"", stream);
-      return;
-    }
-
-  for (i = 0; i < length && things_printed < print_max; ++i)
+  c_string_type result;
+
+  /* We loop because ELTTYPE may be a typedef, and we want to
+     successively peel each typedef until we reach a type we
+     understand.  We don't use CHECK_TYPEDEF because that will strip
+     all typedefs at once -- but in C, wchar_t is itself a typedef, so
+     that would do the wrong thing.  */
+  while (elttype)
     {
     {
-      /* Position of the character we are examining
-         to see whether it is repeated.  */
-      unsigned int rep1;
-      /* Number of repetitions we have detected so far.  */
-      unsigned int reps;
-      unsigned long current_char;
-
-      QUIT;
+      const char *name = TYPE_NAME (elttype);
 
 
-      if (need_comma)
+      if (TYPE_CODE (elttype) == TYPE_CODE_CHAR || !name)
        {
        {
-         fputs_filtered (", ", stream);
-         need_comma = 0;
+         result = C_CHAR;
+         goto done;
        }
 
        }
 
-      current_char = extract_unsigned_integer (string + i * width, width);
+      if (!strcmp (name, "wchar_t"))
+       {
+         result = C_WIDE_CHAR;
+         goto done;
+       }
 
 
-      rep1 = i + 1;
-      reps = 1;
-      while (rep1 < length
-            && extract_unsigned_integer (string + rep1 * width, width)
-            == current_char)
+      if (!strcmp (name, "char16_t"))
        {
        {
-         ++rep1;
-         ++reps;
+         result = C_CHAR_16;
+         goto done;
        }
 
        }
 
-      if (reps > repeat_count_threshold)
+      if (!strcmp (name, "char32_t"))
        {
        {
-         if (in_quotes)
-           {
-             if (inspect_it)
-               fputs_filtered ("\\\", ", stream);
-             else
-               fputs_filtered ("\", ", stream);
-             in_quotes = 0;
-           }
-         LA_PRINT_CHAR (current_char, stream);
-         fprintf_filtered (stream, " <repeats %u times>", reps);
-         i = rep1 - 1;
-         things_printed += repeat_count_threshold;
-         need_comma = 1;
+         result = C_CHAR_32;
+         goto done;
        }
        }
+
+      if (TYPE_CODE (elttype) != TYPE_CODE_TYPEDEF)
+       break;
+
+      /* Call for side effects.  */
+      check_typedef (elttype);
+
+      if (TYPE_TARGET_TYPE (elttype))
+       elttype = TYPE_TARGET_TYPE (elttype);
       else
        {
       else
        {
-         if (!in_quotes)
-           {
-             if (inspect_it)
-               fputs_filtered ("\\\"", stream);
-             else
-               fputs_filtered ("\"", stream);
-             in_quotes = 1;
-           }
-         LA_EMIT_CHAR (current_char, stream, '"');
-         ++things_printed;
+         /* Perhaps check_typedef did not update the target type.  In
+            this case, force the lookup again and hope it works out.
+            It never will for C, but it might for C++.  */
+         elttype = check_typedef (elttype);
        }
     }
 
        }
     }
 
-  /* Terminate the quotes if necessary.  */
-  if (in_quotes)
-    {
-      if (inspect_it)
-       fputs_filtered ("\\\"", stream);
-      else
-       fputs_filtered ("\"", stream);
-    }
+  /* Punt.  */
+  result = C_CHAR;
+
+ done:
+  if (encoding)
+    *encoding = charset_for_string_type (result, gdbarch);
+
+  return result;
+}
+
+/* Print the character C on STREAM as part of the contents of a
+   literal string whose delimiter is QUOTER.  Note that that format
+   for printing characters and strings is language specific.  */
+
+void
+c_emit_char (int c, struct type *type,
+            struct ui_file *stream, int quoter)
+{
+  const char *encoding;
 
 
-  if (force_ellipses || i < length)
-    fputs_filtered ("...", stream);
+  classify_type (type, get_type_arch (type), &encoding);
+  generic_emit_char (c, type, stream, quoter, encoding);
 }
 
 }
 
-/* Create a fundamental C type using default reasonable for the current
-   target machine.
-
-   Some object/debugging file formats (DWARF version 1, COFF, etc) do not
-   define fundamental types such as "int" or "double".  Others (stabs or
-   DWARF version 2, etc) do define fundamental types.  For the formats which
-   don't provide fundamental types, gdb can create such types using this
-   function.
-
-   FIXME:  Some compilers distinguish explicitly signed integral types
-   (signed short, signed int, signed long) from "regular" integral types
-   (short, int, long) in the debugging information.  There is some dis-
-   agreement as to how useful this feature is.  In particular, gcc does
-   not support this.  Also, only some debugging formats allow the
-   distinction to be passed on to a debugger.  For now, we always just
-   use "short", "int", or "long" as the type name, for both the implicit
-   and explicitly signed types.  This also makes life easier for the
-   gdb test suite since we don't have to account for the differences
-   in output depending upon what the compiler and debugging format
-   support.  We will probably have to re-examine the issue when gdb
-   starts taking it's fundamental type information directly from the
-   debugging information supplied by the compiler.  fnf@cygnus.com */
-
-struct type *
-c_create_fundamental_type (struct objfile *objfile, int typeid)
+void
+c_printchar (int c, struct type *type, struct ui_file *stream)
 {
 {
-  struct type *type = NULL;
+  c_string_type str_type;
 
 
-  switch (typeid)
+  str_type = classify_type (type, get_type_arch (type), NULL);
+  switch (str_type)
     {
     {
-    default:
-      /* FIXME:  For now, if we are asked to produce a type not in this
-         language, create the equivalent of a C integer type with the
-         name "<?type?>".  When all the dust settles from the type
-         reconstruction work, this should probably become an error. */
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_INT_BIT / TARGET_CHAR_BIT,
-                       0, "<?type?>", objfile);
-      warning ("internal error: no C/C++ fundamental type %d", typeid);
-      break;
-    case FT_VOID:
-      type = init_type (TYPE_CODE_VOID,
-                       TARGET_CHAR_BIT / TARGET_CHAR_BIT,
-                       0, "void", objfile);
-      break;
-    case FT_BOOLEAN:
-      type = init_type (TYPE_CODE_BOOL,
-                       TARGET_CHAR_BIT / TARGET_CHAR_BIT,
-                       0, "bool", objfile);
-      break;
-    case FT_CHAR:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_CHAR_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_NOSIGN, "char", objfile);
-      break;
-    case FT_SIGNED_CHAR:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_CHAR_BIT / TARGET_CHAR_BIT,
-                       0, "signed char", objfile);
-      break;
-    case FT_UNSIGNED_CHAR:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_CHAR_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_UNSIGNED, "unsigned char", objfile);
-      break;
-    case FT_SHORT:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_SHORT_BIT / TARGET_CHAR_BIT,
-                       0, "short", objfile);
-      break;
-    case FT_SIGNED_SHORT:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_SHORT_BIT / TARGET_CHAR_BIT,
-                       0, "short", objfile);   /* FIXME-fnf */
-      break;
-    case FT_UNSIGNED_SHORT:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_SHORT_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_UNSIGNED, "unsigned short", objfile);
-      break;
-    case FT_INTEGER:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_INT_BIT / TARGET_CHAR_BIT,
-                       0, "int", objfile);
-      break;
-    case FT_SIGNED_INTEGER:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_INT_BIT / TARGET_CHAR_BIT,
-                       0, "int", objfile);     /* FIXME -fnf */
-      break;
-    case FT_UNSIGNED_INTEGER:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_INT_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_UNSIGNED, "unsigned int", objfile);
+    case C_CHAR:
       break;
       break;
-    case FT_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_BIT / TARGET_CHAR_BIT,
-                       0, "long", objfile);
+    case C_WIDE_CHAR:
+      fputc_filtered ('L', stream);
       break;
       break;
-    case FT_SIGNED_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_BIT / TARGET_CHAR_BIT,
-                       0, "long", objfile);    /* FIXME -fnf */
+    case C_CHAR_16:
+      fputc_filtered ('u', stream);
       break;
       break;
-    case FT_UNSIGNED_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_UNSIGNED, "unsigned long", objfile);
+    case C_CHAR_32:
+      fputc_filtered ('U', stream);
       break;
       break;
-    case FT_LONG_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_LONG_BIT / TARGET_CHAR_BIT,
-                       0, "long long", objfile);
-      break;
-    case FT_SIGNED_LONG_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_LONG_BIT / TARGET_CHAR_BIT,
-                       0, "signed long long", objfile);
-      break;
-    case FT_UNSIGNED_LONG_LONG:
-      type = init_type (TYPE_CODE_INT,
-                       TARGET_LONG_LONG_BIT / TARGET_CHAR_BIT,
-                       TYPE_FLAG_UNSIGNED, "unsigned long long", objfile);
-      break;
-    case FT_FLOAT:
-      type = init_type (TYPE_CODE_FLT,
-                       TARGET_FLOAT_BIT / TARGET_CHAR_BIT,
-                       0, "float", objfile);
-      break;
-    case FT_DBL_PREC_FLOAT:
-      type = init_type (TYPE_CODE_FLT,
-                       TARGET_DOUBLE_BIT / TARGET_CHAR_BIT,
-                       0, "double", objfile);
-      break;
-    case FT_EXT_PREC_FLOAT:
-      type = init_type (TYPE_CODE_FLT,
-                       TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT,
-                       0, "long double", objfile);
-      break;
-    case FT_COMPLEX:
-      type = init_type (TYPE_CODE_FLT,
-                       2 * TARGET_FLOAT_BIT / TARGET_CHAR_BIT,
-                       0, "complex float", objfile);
-      TYPE_TARGET_TYPE (type)
-       = init_type (TYPE_CODE_FLT, TARGET_FLOAT_BIT / TARGET_CHAR_BIT,
-                    0, "float", objfile);
+    }
+
+  fputc_filtered ('\'', stream);
+  LA_EMIT_CHAR (c, type, stream, '\'');
+  fputc_filtered ('\'', stream);
+}
+
+/* Print the character string STRING, printing at most LENGTH
+   characters.  LENGTH is -1 if the string is nul terminated.  Each
+   character is WIDTH bytes long.  Printing stops early if the number
+   hits print_max; repeat counts are printed as appropriate.  Print
+   ellipses at the end if we had to stop before printing LENGTH
+   characters, or if FORCE_ELLIPSES.  */
+
+void
+c_printstr (struct ui_file *stream, struct type *type, 
+           const gdb_byte *string, unsigned int length, 
+           const char *user_encoding, int force_ellipses,
+           const struct value_print_options *options)
+{
+  c_string_type str_type;
+  const char *type_encoding;
+  const char *encoding;
+
+  str_type = (classify_type (type, get_type_arch (type), &type_encoding)
+             & ~C_CHAR);
+  switch (str_type)
+    {
+    case C_STRING:
       break;
       break;
-    case FT_DBL_PREC_COMPLEX:
-      type = init_type (TYPE_CODE_FLT,
-                       2 * TARGET_DOUBLE_BIT / TARGET_CHAR_BIT,
-                       0, "complex double", objfile);
-      TYPE_TARGET_TYPE (type)
-       = init_type (TYPE_CODE_FLT, TARGET_DOUBLE_BIT / TARGET_CHAR_BIT,
-                    0, "double", objfile);
+    case C_WIDE_STRING:
+      fputs_filtered ("L", stream);
       break;
       break;
-    case FT_EXT_PREC_COMPLEX:
-      type = init_type (TYPE_CODE_FLT,
-                       2 * TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT,
-                       0, "complex long double", objfile);
-      TYPE_TARGET_TYPE (type)
-       = init_type (TYPE_CODE_FLT, TARGET_LONG_DOUBLE_BIT / TARGET_CHAR_BIT,
-                    0, "long double", objfile);
+    case C_STRING_16:
+      fputs_filtered ("u", stream);
       break;
       break;
-    case FT_TEMPLATE_ARG:
-      type = init_type (TYPE_CODE_TEMPLATE_ARG,
-                       0,
-                       0, "<template arg>", objfile);
+    case C_STRING_32:
+      fputs_filtered ("U", stream);
       break;
     }
       break;
     }
-  return (type);
-}
-\f
-/* Preprocessing and parsing C and C++ expressions.  */
-
 
 
-/* When we find that lexptr (the global var defined in parse.c) is
-   pointing at a macro invocation, we expand the invocation, and call
-   scan_macro_expansion to save the old lexptr here and point lexptr
-   into the expanded text.  When we reach the end of that, we call
-   end_macro_expansion to pop back to the value we saved here.  The
-   macro expansion code promises to return only fully-expanded text,
-   so we don't need to "push" more than one level.
+  encoding = (user_encoding && *user_encoding) ? user_encoding : type_encoding;
 
 
-   This is disgusting, of course.  It would be cleaner to do all macro
-   expansion beforehand, and then hand that to lexptr.  But we don't
-   really know where the expression ends.  Remember, in a command like
+  generic_printstr (stream, type, string, length, encoding, force_ellipses,
+                   '"', 1, options);
+}
 
 
-     (gdb) break *ADDRESS if CONDITION
+/* Obtain a C string from the inferior storing it in a newly allocated
+   buffer in BUFFER, which should be freed by the caller.  If the in-
+   and out-parameter *LENGTH is specified at -1, the string is read
+   until a null character of the appropriate width is found, otherwise
+   the string is read to the length of characters specified.  The size
+   of a character is determined by the length of the target type of
+   the pointer or array.
 
 
-   we evaluate ADDRESS in the scope of the current frame, but we
-   evaluate CONDITION in the scope of the breakpoint's location.  So
-   it's simply wrong to try to macro-expand the whole thing at once.  */
-static char *macro_original_text;
-static char *macro_expanded_text;
+   If VALUE is an array with a known length, and *LENGTH is -1,
+   the function will not read past the end of the array.  However, any
+   declared size of the array is ignored if *LENGTH > 0.
 
 
+   On completion, *LENGTH will be set to the size of the string read in
+   characters.  (If a length of -1 is specified, the length returned
+   will not include the null character).  CHARSET is always set to the
+   target charset.  */
 
 void
 
 void
-scan_macro_expansion (char *expansion)
+c_get_string (struct value *value, gdb::unique_xmalloc_ptr<gdb_byte> *buffer,
+             int *length, struct type **char_type,
+             const char **charset)
 {
 {
-  /* We'd better not be trying to push the stack twice.  */
-  gdb_assert (! macro_original_text);
-  gdb_assert (! macro_expanded_text);
+  int err, width;
+  unsigned int fetchlimit;
+  struct type *type = check_typedef (value_type (value));
+  struct type *element_type = TYPE_TARGET_TYPE (type);
+  int req_length = *length;
+  enum bfd_endian byte_order
+    = type_byte_order (type);
+
+  if (element_type == NULL)
+    goto error;
+
+  if (TYPE_CODE (type) == TYPE_CODE_ARRAY)
+    {
+      /* If we know the size of the array, we can use it as a limit on
+        the number of characters to be fetched.  */
+      if (TYPE_NFIELDS (type) == 1
+         && TYPE_CODE (TYPE_FIELD_TYPE (type, 0)) == TYPE_CODE_RANGE)
+       {
+         LONGEST low_bound, high_bound;
 
 
-  /* Save the old lexptr value, so we can return to it when we're done
-     parsing the expanded text.  */
-  macro_original_text = lexptr;
-  lexptr = expansion;
+         get_discrete_bounds (TYPE_FIELD_TYPE (type, 0),
+                              &low_bound, &high_bound);
+         fetchlimit = high_bound - low_bound + 1;
+       }
+      else
+       fetchlimit = UINT_MAX;
+    }
+  else if (TYPE_CODE (type) == TYPE_CODE_PTR)
+    fetchlimit = UINT_MAX;
+  else
+    /* We work only with arrays and pointers.  */
+    goto error;
+
+  if (! c_textual_element_type (element_type, 0))
+    goto error;
+  classify_type (element_type, get_type_arch (element_type), charset);
+  width = TYPE_LENGTH (element_type);
+
+  /* If the string lives in GDB's memory instead of the inferior's,
+     then we just need to copy it to BUFFER.  Also, since such strings
+     are arrays with known size, FETCHLIMIT will hold the size of the
+     array.
+
+     An array is assumed to live in GDB's memory, so we take this path
+     here.
+
+     However, it's possible for the caller to request more array
+     elements than apparently exist -- this can happen when using the
+     C struct hack.  So, only do this if either no length was
+     specified, or the length is within the existing bounds.  This
+     avoids running off the end of the value's contents.  */
+  if ((VALUE_LVAL (value) == not_lval
+       || VALUE_LVAL (value) == lval_internalvar
+       || TYPE_CODE (type) == TYPE_CODE_ARRAY)
+      && fetchlimit != UINT_MAX
+      && (*length < 0 || *length <= fetchlimit))
+    {
+      int i;
+      const gdb_byte *contents = value_contents (value);
 
 
-  /* Save the expanded text, so we can free it when we're finished.  */
-  macro_expanded_text = expansion;
+      /* If a length is specified, use that.  */
+      if (*length >= 0)
+       i  = *length;
+      else
+       /* Otherwise, look for a null character.  */
+       for (i = 0; i < fetchlimit; i++)
+         if (extract_unsigned_integer (contents + i * width,
+                                       width, byte_order) == 0)
+           break;
+  
+      /* I is now either a user-defined length, the number of non-null
+        characters, or FETCHLIMIT.  */
+      *length = i * width;
+      buffer->reset ((gdb_byte *) xmalloc (*length));
+      memcpy (buffer->get (), contents, *length);
+      err = 0;
+    }
+  else
+    {
+      /* value_as_address does not return an address for an array when
+        c_style_arrays is false, so we handle that specially
+        here.  */
+      CORE_ADDR addr;
+      if (TYPE_CODE (type) == TYPE_CODE_ARRAY)
+       {
+         if (VALUE_LVAL (value) != lval_memory)
+           error (_("Attempt to take address of value "
+                    "not located in memory."));
+         addr = value_address (value);
+       }
+      else
+       addr = value_as_address (value);
+
+      /* Prior to the fix for PR 16196 read_string would ignore fetchlimit
+        if length > 0.  The old "broken" behaviour is the behaviour we want:
+        The caller may want to fetch 100 bytes from a variable length array
+        implemented using the common idiom of having an array of length 1 at
+        the end of a struct.  In this case we want to ignore the declared
+        size of the array.  However, it's counterintuitive to implement that
+        behaviour in read_string: what does fetchlimit otherwise mean if
+        length > 0.  Therefore we implement the behaviour we want here:
+        If *length > 0, don't specify a fetchlimit.  This preserves the
+        previous behaviour.  We could move this check above where we know
+        whether the array is declared with a fixed size, but we only want
+        to apply this behaviour when calling read_string.  PR 16286.  */
+      if (*length > 0)
+       fetchlimit = UINT_MAX;
+
+      err = read_string (addr, *length, width, fetchlimit,
+                        byte_order, buffer, length);
+      if (err != 0)
+       memory_error (TARGET_XFER_E_IO, addr);
+    }
+
+  /* If the LENGTH is specified at -1, we want to return the string
+     length up to the terminating null character.  If an actual length
+     was specified, we want to return the length of exactly what was
+     read.  */
+  if (req_length == -1)
+    /* If the last character is null, subtract it from LENGTH.  */
+    if (*length > 0
+       && extract_unsigned_integer (buffer->get () + *length - width,
+                                    width, byte_order) == 0)
+      *length -= width;
+  
+  /* The read_string function will return the number of bytes read.
+     If length returned from read_string was > 0, return the number of
+     characters read by dividing the number of bytes by width.  */
+  if (*length != 0)
+     *length = *length / width;
+
+  *char_type = element_type;
+
+  return;
+
+ error:
+  {
+    std::string type_str = type_to_string (type);
+    if (!type_str.empty ())
+      {
+       error (_("Trying to read string with inappropriate type `%s'."),
+              type_str.c_str ());
+      }
+    else
+      error (_("Trying to read string with inappropriate type."));
+  }
 }
 
 }
 
+\f
+/* Evaluating C and C++ expressions.  */
+
+/* Convert a UCN.  The digits of the UCN start at P and extend no
+   farther than LIMIT.  DEST_CHARSET is the name of the character set
+   into which the UCN should be converted.  The results are written to
+   OUTPUT.  LENGTH is the maximum length of the UCN, either 4 or 8.
+   Returns a pointer to just after the final digit of the UCN.  */
 
 
-int
-scanning_macro_expansion (void)
+static char *
+convert_ucn (char *p, char *limit, const char *dest_charset,
+            struct obstack *output, int length)
 {
 {
-  return macro_original_text != 0;
+  unsigned long result = 0;
+  gdb_byte data[4];
+  int i;
+
+  for (i = 0; i < length && p < limit && ISXDIGIT (*p); ++i, ++p)
+    result = (result << 4) + host_hex_value (*p);
+
+  for (i = 3; i >= 0; --i)
+    {
+      data[i] = result & 0xff;
+      result >>= 8;
+    }
+
+  convert_between_encodings ("UTF-32BE", dest_charset, data,
+                            4, 4, output, translit_none);
+
+  return p;
 }
 
 }
 
+/* Emit a character, VALUE, which was specified numerically, to
+   OUTPUT.  TYPE is the target character type.  */
 
 
-void 
-finished_macro_expansion (void)
+static void
+emit_numeric_character (struct type *type, unsigned long value,
+                       struct obstack *output)
 {
 {
-  /* There'd better be something to pop back to, and we better have
-     saved a pointer to the start of the expanded text.  */
-  gdb_assert (macro_original_text);
-  gdb_assert (macro_expanded_text);
-
-  /* Pop back to the original text.  */
-  lexptr = macro_original_text;
-  macro_original_text = 0;
-
-  /* Free the expanded text.  */
-  xfree (macro_expanded_text);
-  macro_expanded_text = 0;
+  gdb_byte *buffer;
+
+  buffer = (gdb_byte *) alloca (TYPE_LENGTH (type));
+  pack_long (buffer, type, value);
+  obstack_grow (output, buffer, TYPE_LENGTH (type));
 }
 
 }
 
+/* Convert an octal escape sequence.  TYPE is the target character
+   type.  The digits of the escape sequence begin at P and extend no
+   farther than LIMIT.  The result is written to OUTPUT.  Returns a
+   pointer to just after the final digit of the escape sequence.  */
 
 
-static void
-scan_macro_cleanup (void *dummy)
+static char *
+convert_octal (struct type *type, char *p, 
+              char *limit, struct obstack *output)
 {
 {
-  if (macro_original_text)
-    finished_macro_expansion ();
+  int i;
+  unsigned long value = 0;
+
+  for (i = 0;
+       i < 3 && p < limit && ISDIGIT (*p) && *p != '8' && *p != '9';
+       ++i)
+    {
+      value = 8 * value + host_hex_value (*p);
+      ++p;
+    }
+
+  emit_numeric_character (type, value, output);
+
+  return p;
 }
 
 }
 
+/* Convert a hex escape sequence.  TYPE is the target character type.
+   The digits of the escape sequence begin at P and extend no farther
+   than LIMIT.  The result is written to OUTPUT.  Returns a pointer to
+   just after the final digit of the escape sequence.  */
+
+static char *
+convert_hex (struct type *type, char *p,
+            char *limit, struct obstack *output)
+{
+  unsigned long value = 0;
+
+  while (p < limit && ISXDIGIT (*p))
+    {
+      value = 16 * value + host_hex_value (*p);
+      ++p;
+    }
 
 
-/* We set these global variables before calling c_parse, to tell it
-   how it to find macro definitions for the expression at hand.  */
-macro_lookup_ftype *expression_macro_lookup_func;
-void *expression_macro_lookup_baton;
+  emit_numeric_character (type, value, output);
 
 
+  return p;
+}
 
 
-static struct macro_definition *
-null_macro_lookup (const char *name, void *baton)
+#define ADVANCE                                        \
+  do {                                         \
+    ++p;                                       \
+    if (p == limit)                            \
+      error (_("Malformed escape sequence"));  \
+  } while (0)
+
+/* Convert an escape sequence to a target format.  TYPE is the target
+   character type to use, and DEST_CHARSET is the name of the target
+   character set.  The backslash of the escape sequence is at *P, and
+   the escape sequence will not extend past LIMIT.  The results are
+   written to OUTPUT.  Returns a pointer to just past the final
+   character of the escape sequence.  */
+
+static char *
+convert_escape (struct type *type, const char *dest_charset,
+               char *p, char *limit, struct obstack *output)
 {
 {
-  return 0;
+  /* Skip the backslash.  */
+  ADVANCE;
+
+  switch (*p)
+    {
+    case '\\':
+      obstack_1grow (output, '\\');
+      ++p;
+      break;
+
+    case 'x':
+      ADVANCE;
+      if (!ISXDIGIT (*p))
+       error (_("\\x used with no following hex digits."));
+      p = convert_hex (type, p, limit, output);
+      break;
+
+    case '0':
+    case '1':
+    case '2':
+    case '3':
+    case '4':
+    case '5':
+    case '6':
+    case '7':
+      p = convert_octal (type, p, limit, output);
+      break;
+
+    case 'u':
+    case 'U':
+      {
+       int length = *p == 'u' ? 4 : 8;
+
+       ADVANCE;
+       if (!ISXDIGIT (*p))
+         error (_("\\u used with no following hex digits"));
+       p = convert_ucn (p, limit, dest_charset, output, length);
+      }
+    }
+
+  return p;
 }
 
 }
 
+/* Given a single string from a (C-specific) OP_STRING list, convert
+   it to a target string, handling escape sequences specially.  The
+   output is written to OUTPUT.  DATA is the input string, which has
+   length LEN.  DEST_CHARSET is the name of the target character set,
+   and TYPE is the type of target character to use.  */
 
 
-static int
-c_preprocess_and_parse (void)
+static void
+parse_one_string (struct obstack *output, char *data, int len,
+                 const char *dest_charset, struct type *type)
 {
 {
-  /* Set up a lookup function for the macro expander.  */
-  struct macro_scope *scope = 0;
-  struct cleanup *back_to = make_cleanup (free_current_contents, &scope);
+  char *limit;
 
 
-  if (expression_context_block)
-    scope = sal_macro_scope (find_pc_line (expression_context_pc, 0));
-  else
-    scope = default_macro_scope ();
+  limit = data + len;
 
 
-  if (scope)
+  while (data < limit)
     {
     {
-      expression_macro_lookup_func = standard_macro_lookup;
-      expression_macro_lookup_baton = (void *) scope;
+      char *p = data;
+
+      /* Look for next escape, or the end of the input.  */
+      while (p < limit && *p != '\\')
+       ++p;
+      /* If we saw a run of characters, convert them all.  */
+      if (p > data)
+       convert_between_encodings (host_charset (), dest_charset,
+                                  (gdb_byte *) data, p - data, 1,
+                                  output, translit_none);
+      /* If we saw an escape, convert it.  */
+      if (p < limit)
+       p = convert_escape (type, dest_charset, p, limit, output);
+      data = p;
     }
     }
-  else
+}
+
+/* Expression evaluator for the C language family.  Most operations
+   are delegated to evaluate_subexp_standard; see that function for a
+   description of the arguments.  */
+
+struct value *
+evaluate_subexp_c (struct type *expect_type, struct expression *exp,
+                  int *pos, enum noside noside)
+{
+  enum exp_opcode op = exp->elts[*pos].opcode;
+
+  switch (op)
     {
     {
-      expression_macro_lookup_func = null_macro_lookup;
-      expression_macro_lookup_baton = 0;      
-    }
+    case OP_STRING:
+      {
+       int oplen, limit;
+       struct type *type;
+       struct value *result;
+       c_string_type dest_type;
+       const char *dest_charset;
+       int satisfy_expected = 0;
+
+       auto_obstack output;
+
+       ++*pos;
+       oplen = longest_to_int (exp->elts[*pos].longconst);
+
+       ++*pos;
+       limit = *pos + BYTES_TO_EXP_ELEM (oplen + 1);
+       dest_type = ((enum c_string_type_values)
+                    longest_to_int (exp->elts[*pos].longconst));
+       switch (dest_type & ~C_CHAR)
+         {
+         case C_STRING:
+           type = language_string_char_type (exp->language_defn,
+                                             exp->gdbarch);
+           break;
+         case C_WIDE_STRING:
+           type = lookup_typename (exp->language_defn, "wchar_t", NULL, 0);
+           break;
+         case C_STRING_16:
+           type = lookup_typename (exp->language_defn, "char16_t", NULL, 0);
+           break;
+         case C_STRING_32:
+           type = lookup_typename (exp->language_defn, "char32_t", NULL, 0);
+           break;
+         default:
+           internal_error (__FILE__, __LINE__, _("unhandled c_string_type"));
+         }
+
+       /* Ensure TYPE_LENGTH is valid for TYPE.  */
+       check_typedef (type);
+
+       /* If the caller expects an array of some integral type,
+          satisfy them.  If something odder is expected, rely on the
+          caller to cast.  */
+       if (expect_type && TYPE_CODE (expect_type) == TYPE_CODE_ARRAY)
+         {
+           struct type *element_type
+             = check_typedef (TYPE_TARGET_TYPE (expect_type));
+
+           if (TYPE_CODE (element_type) == TYPE_CODE_INT
+               || TYPE_CODE (element_type) == TYPE_CODE_CHAR)
+             {
+               type = element_type;
+               satisfy_expected = 1;
+             }
+         }
+
+       dest_charset = charset_for_string_type (dest_type, exp->gdbarch);
+
+       ++*pos;
+       while (*pos < limit)
+         {
+           int len;
+
+           len = longest_to_int (exp->elts[*pos].longconst);
+
+           ++*pos;
+           if (noside != EVAL_SKIP)
+             parse_one_string (&output, &exp->elts[*pos].string, len,
+                               dest_charset, type);
+           *pos += BYTES_TO_EXP_ELEM (len);
+         }
+
+       /* Skip the trailing length and opcode.  */
+       *pos += 2;
+
+       if (noside == EVAL_SKIP)
+         {
+           /* Return a dummy value of the appropriate type.  */
+           if (expect_type != NULL)
+             result = allocate_value (expect_type);
+           else if ((dest_type & C_CHAR) != 0)
+             result = allocate_value (type);
+           else
+             result = value_cstring ("", 0, type);
+           return result;
+         }
+
+       if ((dest_type & C_CHAR) != 0)
+         {
+           LONGEST value;
+
+           if (obstack_object_size (&output) != TYPE_LENGTH (type))
+             error (_("Could not convert character "
+                      "constant to target character set"));
+           value = unpack_long (type, (gdb_byte *) obstack_base (&output));
+           result = value_from_longest (type, value);
+         }
+       else
+         {
+           int i;
+
+           /* Write the terminating character.  */
+           for (i = 0; i < TYPE_LENGTH (type); ++i)
+             obstack_1grow (&output, 0);
+
+           if (satisfy_expected)
+             {
+               LONGEST low_bound, high_bound;
+               int element_size = TYPE_LENGTH (type);
+
+               if (get_discrete_bounds (TYPE_INDEX_TYPE (expect_type),
+                                        &low_bound, &high_bound) < 0)
+                 {
+                   low_bound = 0;
+                   high_bound = (TYPE_LENGTH (expect_type) / element_size) - 1;
+                 }
+               if (obstack_object_size (&output) / element_size
+                   > (high_bound - low_bound + 1))
+                 error (_("Too many array elements"));
+
+               result = allocate_value (expect_type);
+               memcpy (value_contents_raw (result), obstack_base (&output),
+                       obstack_object_size (&output));
+             }
+           else
+             result = value_cstring ((const char *) obstack_base (&output),
+                                     obstack_object_size (&output),
+                                     type);
+         }
+       return result;
+      }
+      break;
 
 
-  gdb_assert (! macro_original_text);
-  make_cleanup (scan_macro_cleanup, 0);
+    default:
+      break;
+    }
+  return evaluate_subexp_standard (expect_type, exp, pos, noside);
+}
+\f
+/* la_watch_location_expression for C.  */
 
 
-  {
-    int result = c_parse ();
-    do_cleanups (back_to);
-    return result;
-  }
+gdb::unique_xmalloc_ptr<char>
+c_watch_location_expression (struct type *type, CORE_ADDR addr)
+{
+  type = check_typedef (TYPE_TARGET_TYPE (check_typedef (type)));
+  std::string name = type_to_string (type);
+  return gdb::unique_xmalloc_ptr<char>
+    (xstrprintf ("* (%s *) %s", name.c_str (), core_addr_to_string (addr)));
 }
 
 }
 
+/* See c-lang.h.  */
+
+bool
+c_is_string_type_p (struct type *type)
+{
+  type = check_typedef (type);
+  while (TYPE_CODE (type) == TYPE_CODE_REF)
+    {
+      type = TYPE_TARGET_TYPE (type);
+      type = check_typedef (type);
+    }
+
+  switch (TYPE_CODE (type))
+    {
+    case TYPE_CODE_ARRAY:
+      {
+       /* See if target type looks like a string.  */
+       struct type *array_target_type = TYPE_TARGET_TYPE (type);
+       return (TYPE_LENGTH (type) > 0
+               && TYPE_LENGTH (array_target_type) > 0
+               && c_textual_element_type (array_target_type, 0));
+      }
+    case TYPE_CODE_STRING:
+      return true;
+    case TYPE_CODE_PTR:
+      {
+       struct type *element_type = TYPE_TARGET_TYPE (type);
+       return c_textual_element_type (element_type, 0);
+      }
+    default:
+      break;
+    }
+
+  return false;
+}
 
 \f
 /* Table mapping opcodes into strings for printing operators
 
 \f
 /* Table mapping opcodes into strings for printing operators
@@ -502,15 +799,17 @@ const struct op_print c_op_print_tab[] =
   {"/", BINOP_DIV, PREC_MUL, 0},
   {"%", BINOP_REM, PREC_MUL, 0},
   {"@", BINOP_REPEAT, PREC_REPEAT, 0},
   {"/", BINOP_DIV, PREC_MUL, 0},
   {"%", BINOP_REM, PREC_MUL, 0},
   {"@", BINOP_REPEAT, PREC_REPEAT, 0},
+  {"+", UNOP_PLUS, PREC_PREFIX, 0},
   {"-", UNOP_NEG, PREC_PREFIX, 0},
   {"!", UNOP_LOGICAL_NOT, PREC_PREFIX, 0},
   {"~", UNOP_COMPLEMENT, PREC_PREFIX, 0},
   {"*", UNOP_IND, PREC_PREFIX, 0},
   {"&", UNOP_ADDR, PREC_PREFIX, 0},
   {"sizeof ", UNOP_SIZEOF, PREC_PREFIX, 0},
   {"-", UNOP_NEG, PREC_PREFIX, 0},
   {"!", UNOP_LOGICAL_NOT, PREC_PREFIX, 0},
   {"~", UNOP_COMPLEMENT, PREC_PREFIX, 0},
   {"*", UNOP_IND, PREC_PREFIX, 0},
   {"&", UNOP_ADDR, PREC_PREFIX, 0},
   {"sizeof ", UNOP_SIZEOF, PREC_PREFIX, 0},
+  {"alignof ", UNOP_ALIGNOF, PREC_PREFIX, 0},
   {"++", UNOP_PREINCREMENT, PREC_PREFIX, 0},
   {"--", UNOP_PREDECREMENT, PREC_PREFIX, 0},
   {"++", UNOP_PREINCREMENT, PREC_PREFIX, 0},
   {"--", UNOP_PREDECREMENT, PREC_PREFIX, 0},
-  {NULL, 0, 0, 0}
+  {NULL, OP_NULL, PREC_PREFIX, 0}
 };
 \f
 enum c_primitive_types {
 };
 \f
 enum c_primitive_types {
@@ -531,6 +830,9 @@ enum c_primitive_types {
   c_primitive_type_long_double,
   c_primitive_type_complex,
   c_primitive_type_double_complex,
   c_primitive_type_long_double,
   c_primitive_type_complex,
   c_primitive_type_double_complex,
+  c_primitive_type_decfloat,
+  c_primitive_type_decdouble,
+  c_primitive_type_declong,
   nr_c_primitive_types
 };
 
   nr_c_primitive_types
 };
 
@@ -539,6 +841,7 @@ c_language_arch_info (struct gdbarch *gdbarch,
                      struct language_arch_info *lai)
 {
   const struct builtin_type *builtin = builtin_type (gdbarch);
                      struct language_arch_info *lai)
 {
   const struct builtin_type *builtin = builtin_type (gdbarch);
+
   lai->string_char_type = builtin->builtin_char;
   lai->primitive_type_vector
     = GDBARCH_OBSTACK_CALLOC (gdbarch, nr_c_primitive_types + 1,
   lai->string_char_type = builtin->builtin_char;
   lai->primitive_type_vector
     = GDBARCH_OBSTACK_CALLOC (gdbarch, nr_c_primitive_types + 1,
@@ -560,146 +863,274 @@ c_language_arch_info (struct gdbarch *gdbarch,
   lai->primitive_type_vector [c_primitive_type_long_double] = builtin->builtin_long_double;
   lai->primitive_type_vector [c_primitive_type_complex] = builtin->builtin_complex;
   lai->primitive_type_vector [c_primitive_type_double_complex] = builtin->builtin_double_complex;
   lai->primitive_type_vector [c_primitive_type_long_double] = builtin->builtin_long_double;
   lai->primitive_type_vector [c_primitive_type_complex] = builtin->builtin_complex;
   lai->primitive_type_vector [c_primitive_type_double_complex] = builtin->builtin_double_complex;
+  lai->primitive_type_vector [c_primitive_type_decfloat] = builtin->builtin_decfloat;
+  lai->primitive_type_vector [c_primitive_type_decdouble] = builtin->builtin_decdouble;
+  lai->primitive_type_vector [c_primitive_type_declong] = builtin->builtin_declong;
+
+  lai->bool_type_default = builtin->builtin_int;
+}
+
+const struct exp_descriptor exp_descriptor_c = 
+{
+  print_subexp_standard,
+  operator_length_standard,
+  operator_check_standard,
+  op_name_standard,
+  dump_subexp_body_standard,
+  evaluate_subexp_c
+};
+
+static const char *c_extensions[] =
+{
+  ".c", NULL
 };
 
 };
 
-const struct language_defn c_language_defn =
+extern const struct language_defn c_language_defn =
 {
   "c",                         /* Language name */
 {
   "c",                         /* Language name */
+  "C",
   language_c,
   language_c,
-  NULL,
   range_check_off,
   range_check_off,
-  type_check_off,
   case_sensitive_on,
   array_row_major,
   case_sensitive_on,
   array_row_major,
-  &exp_descriptor_standard,
-  c_preprocess_and_parse,
-  c_error,
+  macro_expansion_c,
+  c_extensions,
+  &exp_descriptor_c,
+  c_parse,
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
-  c_create_fundamental_type,   /* Create fundamental type in this language */
   c_print_type,                        /* Print a type using appropriate syntax */
   c_print_type,                        /* Print a type using appropriate syntax */
+  c_print_typedef,             /* Print a typedef using appropriate syntax */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
+  default_read_var_value,      /* la_read_var_value */
   NULL,                                /* Language specific skip_trampoline */
   NULL,                                /* Language specific skip_trampoline */
-  NULL,                                /* value_of_this */
+  NULL,                                /* name_of_this */
+  true,                                /* la_store_sym_names_in_linkage_form_p */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
-  NULL,                                /* Language specific class_name_from_physname */
-  {"", "", "", ""},            /* Binary format info */
-  {"0%lo", "0", "o", ""},      /* Octal format info */
-  {"%ld", "", "d", ""},                /* Decimal format info */
-  {"0x%lx", "0x", "x", ""},    /* Hex format info */
+  NULL,
+  NULL,                                /* Language specific
+                                  class_name_from_physname */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
-  NULL,
   default_word_break_characters,
   default_word_break_characters,
+  default_collect_symbol_completion_matches,
   c_language_arch_info,
   c_language_arch_info,
-  LANG_MAGIC
+  default_print_array_index,
+  default_pass_by_reference,
+  c_watch_location_expression,
+  NULL,                                /* la_get_symbol_name_matcher */
+  iterate_over_symbols,
+  default_search_name_hash,
+  &c_varobj_ops,
+  c_get_compile_context,
+  c_compute_program,
+  c_is_string_type_p,
+  "{...}"                      /* la_struct_too_deep_ellipsis */
+};
+
+enum cplus_primitive_types {
+  cplus_primitive_type_int,
+  cplus_primitive_type_long,
+  cplus_primitive_type_short,
+  cplus_primitive_type_char,
+  cplus_primitive_type_float,
+  cplus_primitive_type_double,
+  cplus_primitive_type_void,
+  cplus_primitive_type_long_long,
+  cplus_primitive_type_signed_char,
+  cplus_primitive_type_unsigned_char,
+  cplus_primitive_type_unsigned_short,
+  cplus_primitive_type_unsigned_int,
+  cplus_primitive_type_unsigned_long,
+  cplus_primitive_type_unsigned_long_long,
+  cplus_primitive_type_long_double,
+  cplus_primitive_type_complex,
+  cplus_primitive_type_double_complex,
+  cplus_primitive_type_bool,
+  cplus_primitive_type_decfloat,
+  cplus_primitive_type_decdouble,
+  cplus_primitive_type_declong,
+  cplus_primitive_type_char16_t,
+  cplus_primitive_type_char32_t,
+  cplus_primitive_type_wchar_t,
+  nr_cplus_primitive_types
 };
 
 };
 
-struct type **const (cplus_builtin_types[]) =
+static void
+cplus_language_arch_info (struct gdbarch *gdbarch,
+                         struct language_arch_info *lai)
 {
 {
-  &builtin_type_int,
-  &builtin_type_long,
-  &builtin_type_short,
-  &builtin_type_char,
-  &builtin_type_float,
-  &builtin_type_double,
-  &builtin_type_void,
-  &builtin_type_long_long,
-  &builtin_type_signed_char,
-  &builtin_type_unsigned_char,
-  &builtin_type_unsigned_short,
-  &builtin_type_unsigned_int,
-  &builtin_type_unsigned_long,
-  &builtin_type_unsigned_long_long,
-  &builtin_type_long_double,
-  &builtin_type_complex,
-  &builtin_type_double_complex,
-  &builtin_type_bool,
-  0
+  const struct builtin_type *builtin = builtin_type (gdbarch);
+
+  lai->string_char_type = builtin->builtin_char;
+  lai->primitive_type_vector
+    = GDBARCH_OBSTACK_CALLOC (gdbarch, nr_cplus_primitive_types + 1,
+                             struct type *);
+  lai->primitive_type_vector [cplus_primitive_type_int]
+    = builtin->builtin_int;
+  lai->primitive_type_vector [cplus_primitive_type_long]
+    = builtin->builtin_long;
+  lai->primitive_type_vector [cplus_primitive_type_short]
+    = builtin->builtin_short;
+  lai->primitive_type_vector [cplus_primitive_type_char]
+    = builtin->builtin_char;
+  lai->primitive_type_vector [cplus_primitive_type_float]
+    = builtin->builtin_float;
+  lai->primitive_type_vector [cplus_primitive_type_double]
+    = builtin->builtin_double;
+  lai->primitive_type_vector [cplus_primitive_type_void]
+    = builtin->builtin_void;
+  lai->primitive_type_vector [cplus_primitive_type_long_long]
+    = builtin->builtin_long_long;
+  lai->primitive_type_vector [cplus_primitive_type_signed_char]
+    = builtin->builtin_signed_char;
+  lai->primitive_type_vector [cplus_primitive_type_unsigned_char]
+    = builtin->builtin_unsigned_char;
+  lai->primitive_type_vector [cplus_primitive_type_unsigned_short]
+    = builtin->builtin_unsigned_short;
+  lai->primitive_type_vector [cplus_primitive_type_unsigned_int]
+    = builtin->builtin_unsigned_int;
+  lai->primitive_type_vector [cplus_primitive_type_unsigned_long]
+    = builtin->builtin_unsigned_long;
+  lai->primitive_type_vector [cplus_primitive_type_unsigned_long_long]
+    = builtin->builtin_unsigned_long_long;
+  lai->primitive_type_vector [cplus_primitive_type_long_double]
+    = builtin->builtin_long_double;
+  lai->primitive_type_vector [cplus_primitive_type_complex]
+    = builtin->builtin_complex;
+  lai->primitive_type_vector [cplus_primitive_type_double_complex]
+    = builtin->builtin_double_complex;
+  lai->primitive_type_vector [cplus_primitive_type_bool]
+    = builtin->builtin_bool;
+  lai->primitive_type_vector [cplus_primitive_type_decfloat]
+    = builtin->builtin_decfloat;
+  lai->primitive_type_vector [cplus_primitive_type_decdouble]
+    = builtin->builtin_decdouble;
+  lai->primitive_type_vector [cplus_primitive_type_declong]
+    = builtin->builtin_declong;
+  lai->primitive_type_vector [cplus_primitive_type_char16_t]
+    = builtin->builtin_char16;
+  lai->primitive_type_vector [cplus_primitive_type_char32_t]
+    = builtin->builtin_char32;
+  lai->primitive_type_vector [cplus_primitive_type_wchar_t]
+    = builtin->builtin_wchar;
+
+  lai->bool_type_symbol = "bool";
+  lai->bool_type_default = builtin->builtin_bool;
+}
+
+static const char *cplus_extensions[] =
+{
+  ".C", ".cc", ".cp", ".cpp", ".cxx", ".c++", NULL
 };
 
 };
 
-const struct language_defn cplus_language_defn =
+extern const struct language_defn cplus_language_defn =
 {
   "c++",                       /* Language name */
 {
   "c++",                       /* Language name */
+  "C++",
   language_cplus,
   language_cplus,
-  cplus_builtin_types,
   range_check_off,
   range_check_off,
-  type_check_off,
   case_sensitive_on,
   array_row_major,
   case_sensitive_on,
   array_row_major,
-  &exp_descriptor_standard,
-  c_preprocess_and_parse,
-  c_error,
+  macro_expansion_c,
+  cplus_extensions,
+  &exp_descriptor_c,
+  c_parse,
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
-  c_create_fundamental_type,   /* Create fundamental type in this language */
   c_print_type,                        /* Print a type using appropriate syntax */
   c_print_type,                        /* Print a type using appropriate syntax */
+  c_print_typedef,             /* Print a typedef using appropriate syntax */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
-  NULL,                                /* Language specific skip_trampoline */
-  value_of_this,               /* value_of_this */
+  default_read_var_value,      /* la_read_var_value */
+  cplus_skip_trampoline,       /* Language specific skip_trampoline */
+  "this",                       /* name_of_this */
+  false,                       /* la_store_sym_names_in_linkage_form_p */
   cp_lookup_symbol_nonlocal,   /* lookup_symbol_nonlocal */
   cp_lookup_transparent_type,   /* lookup_transparent_type */
   cp_lookup_symbol_nonlocal,   /* lookup_symbol_nonlocal */
   cp_lookup_transparent_type,   /* lookup_transparent_type */
-  cplus_demangle,              /* Language specific symbol demangler */
-  cp_class_name_from_physname,  /* Language specific class_name_from_physname */
-  {"", "", "", ""},            /* Binary format info */
-  {"0%lo", "0", "o", ""},      /* Octal format info */
-  {"%ld", "", "d", ""},                /* Decimal format info */
-  {"0x%lx", "0x", "x", ""},    /* Hex format info */
+  gdb_demangle,                        /* Language specific symbol demangler */
+  gdb_sniff_from_mangled_name,
+  cp_class_name_from_physname,  /* Language specific
+                                  class_name_from_physname */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
-  &builtin_type_char,          /* Type of string elements */
   default_word_break_characters,
   default_word_break_characters,
-  NULL, /* FIXME: la_language_arch_info.  */
-  LANG_MAGIC
+  default_collect_symbol_completion_matches,
+  cplus_language_arch_info,
+  default_print_array_index,
+  cp_pass_by_reference,
+  c_watch_location_expression,
+  cp_get_symbol_name_matcher,
+  iterate_over_symbols,
+  cp_search_name_hash,
+  &cplus_varobj_ops,
+  cplus_get_compile_context,
+  cplus_compute_program,
+  c_is_string_type_p,
+  "{...}"                      /* la_struct_too_deep_ellipsis */
+};
+
+static const char *asm_extensions[] =
+{
+  ".s", ".sx", ".S", NULL
 };
 
 };
 
-const struct language_defn asm_language_defn =
+extern const struct language_defn asm_language_defn =
 {
   "asm",                       /* Language name */
 {
   "asm",                       /* Language name */
+  "assembly",
   language_asm,
   language_asm,
-  NULL,
   range_check_off,
   range_check_off,
-  type_check_off,
   case_sensitive_on,
   array_row_major,
   case_sensitive_on,
   array_row_major,
-  &exp_descriptor_standard,
-  c_preprocess_and_parse,
-  c_error,
+  macro_expansion_c,
+  asm_extensions,
+  &exp_descriptor_c,
+  c_parse,
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
-  c_create_fundamental_type,   /* Create fundamental type in this language */
   c_print_type,                        /* Print a type using appropriate syntax */
   c_print_type,                        /* Print a type using appropriate syntax */
+  c_print_typedef,             /* Print a typedef using appropriate syntax */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
+  default_read_var_value,      /* la_read_var_value */
   NULL,                                /* Language specific skip_trampoline */
   NULL,                                /* Language specific skip_trampoline */
-  NULL,                                /* value_of_this */
+  NULL,                                /* name_of_this */
+  true,                                /* la_store_sym_names_in_linkage_form_p */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
-  NULL,                                /* Language specific class_name_from_physname */
-  {"", "", "", ""},            /* Binary format info */
-  {"0%lo", "0", "o", ""},      /* Octal format info */
-  {"%ld", "", "d", ""},                /* Decimal format info */
-  {"0x%lx", "0x", "x", ""},    /* Hex format info */
+  NULL,
+  NULL,                                /* Language specific
+                                  class_name_from_physname */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
-  NULL,
   default_word_break_characters,
   default_word_break_characters,
-  c_language_arch_info, /* FIXME: la_language_arch_info.  */
-  LANG_MAGIC
+  default_collect_symbol_completion_matches,
+  c_language_arch_info,                /* FIXME: la_language_arch_info.  */
+  default_print_array_index,
+  default_pass_by_reference,
+  c_watch_location_expression,
+  NULL,                                /* la_get_symbol_name_matcher */
+  iterate_over_symbols,
+  default_search_name_hash,
+  &default_varobj_ops,
+  NULL,
+  NULL,
+  c_is_string_type_p,
+  "{...}"                      /* la_struct_too_deep_ellipsis */
 };
 
 /* The following language_defn does not represent a real language.
 };
 
 /* The following language_defn does not represent a real language.
@@ -707,50 +1138,51 @@ const struct language_defn asm_language_defn =
    to do some simple operations when debugging applications that use
    a language currently not supported by GDB.  */
 
    to do some simple operations when debugging applications that use
    a language currently not supported by GDB.  */
 
-const struct language_defn minimal_language_defn =
+extern const struct language_defn minimal_language_defn =
 {
   "minimal",                   /* Language name */
 {
   "minimal",                   /* Language name */
+  "Minimal",
   language_minimal,
   language_minimal,
-  NULL,
   range_check_off,
   range_check_off,
-  type_check_off,
   case_sensitive_on,
   array_row_major,
   case_sensitive_on,
   array_row_major,
-  &exp_descriptor_standard,
-  c_preprocess_and_parse,
-  c_error,
+  macro_expansion_c,
+  NULL,
+  &exp_descriptor_c,
+  c_parse,
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
   null_post_parser,
   c_printchar,                 /* Print a character constant */
   c_printstr,                  /* Function to print string constant */
   c_emit_char,                 /* Print a single char */
-  c_create_fundamental_type,   /* Create fundamental type in this language */
   c_print_type,                        /* Print a type using appropriate syntax */
   c_print_type,                        /* Print a type using appropriate syntax */
+  c_print_typedef,             /* Print a typedef using appropriate syntax */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
   c_val_print,                 /* Print a value using appropriate syntax */
   c_value_print,               /* Print a top-level value */
+  default_read_var_value,      /* la_read_var_value */
   NULL,                                /* Language specific skip_trampoline */
   NULL,                                /* Language specific skip_trampoline */
-  NULL,                                /* value_of_this */
+  NULL,                                /* name_of_this */
+  true,                                /* la_store_sym_names_in_linkage_form_p */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
   basic_lookup_symbol_nonlocal,        /* lookup_symbol_nonlocal */
   basic_lookup_transparent_type,/* lookup_transparent_type */
   NULL,                                /* Language specific symbol demangler */
-  NULL,                                /* Language specific class_name_from_physname */
-  {"", "", "", ""},            /* Binary format info */
-  {"0%lo", "0", "o", ""},      /* Octal format info */
-  {"%ld", "", "d", ""},                /* Decimal format info */
-  {"0x%lx", "0x", "x", ""},    /* Hex format info */
+  NULL,
+  NULL,                                /* Language specific
+                                  class_name_from_physname */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
   c_op_print_tab,              /* expression operators for printing */
   1,                           /* c-style arrays */
   0,                           /* String lower bound */
-  NULL,
   default_word_break_characters,
   default_word_break_characters,
+  default_collect_symbol_completion_matches,
   c_language_arch_info,
   c_language_arch_info,
-  LANG_MAGIC
+  default_print_array_index,
+  default_pass_by_reference,
+  c_watch_location_expression,
+  NULL,                                /* la_get_symbol_name_matcher */
+  iterate_over_symbols,
+  default_search_name_hash,
+  &default_varobj_ops,
+  NULL,
+  NULL,
+  c_is_string_type_p,
+  "{...}"                      /* la_struct_too_deep_ellipsis */
 };
 };
-
-void
-_initialize_c_language (void)
-{
-  add_language (&c_language_defn);
-  add_language (&cplus_language_defn);
-  add_language (&asm_language_defn);
-  add_language (&minimal_language_defn);
-}
This page took 0.039895 seconds and 4 git commands to generate.