______________________________________________________________________

  18   Language support library         [lib.language.support]

  ______________________________________________________________________

1 This clause describes the function signatures that are called  implic­
  itly, and the types of objects generated implicitly, during the execu­
  tion of some C++ programs.  It also describes the headers that declare
  these function signatures and define any related types.

2 The   following  subclauses  describe  common  type  definitions  used
  throughout the library, characteristics of the predefined types, func­
  tions  supporting  start and termination of a C++ program, support for
  dynamic memory management, support for  dynamic  type  identification,
  support for exception processing, and other runtime support, as summa­
  rized in Table 1:

                Table 1--Language support library summary

     +--------------------------------------------------------------+
     |                   Subclause                       Header(s)  |
     +--------------------------------------------------------------+
     |_lib.support.types_ Types                         <cstddef>   |
     +--------------------------------------------------------------+
     |                                                  <limits>    |
     |_lib.support.limits_ Implementation properties    <climits>   |
     |                                                  <cfloat>    |
     +--------------------------------------------------------------+
     |_lib.support.start.term_ Start and termination    <cstdlib>   |
     +--------------------------------------------------------------+
     |_lib.support.dynamic_ Dynamic memory management   <new>       |
     +--------------------------------------------------------------+
     |_lib.support.rtti_ Type identification            <typeinfo>  |
     +--------------------------------------------------------------+
     |_lib.support.exception_ Exception handling        <exception> |
     +--------------------------------------------------------------+
     |                                                  <cstdarg>   |
     |                                                  <csetjmp>   |
     |_lib.support.runtime_ Other runtime support       <ctime>     |
     |                                                  <csignal>   |
     |                                                  <cstdlib>   |
     +--------------------------------------------------------------+

  18.1  Types                                        [lib.support.types]

1 Common definitons.

2 Header <cstddef> (Table 2):

                    Table 2--Header <cstddef> synopsis

          +----------------------------------------------------+
          | Type                      Name(s)                  |
          +----------------------------------------------------+
          |Macros:   NULL <cstddef>       offsetof             |
          +----------------------------------------------------+
          |Types:    ptrdiff_t<cstddef>   size_t <cstddef>     |
          +----------------------------------------------------+

3 The contents are the same as the Standard C library, with the  follow­
  ing changes:

4 The  macro NULL is an implementation-defined C++ null-pointer constant
  in this International Standard (_conv.ptr_).1)

5 The macro offsetof accepts a restricted set of type arguments in  this
  International  Standard.  type shall be a POD structure or a POD union
  (_class_).

  SEE ALSO: subclause _expr.sizeof_, Sizeof, subclause _expr.add_, Addi­
  tive  operators,  subclause   _class.free_, Free store, and ISO C sub­
  clause 7.1.6.

  18.2  Implementation properties                   [lib.support.limits]

1 Characteristics   of   implementation-dependent   fundamental    types
  (_basic.fundamental_).

  18.2.1  Numeric limits                                    [lib.limits]

1 The  numeric_limits  component provides a C++ program with information
  about various properties of the implementation's representation of the
  fundamental types.

2 Specializations  shall  be  provided  for  each fundamental type, both
  floating point and integer, including bool.  The member is_specialized
  shall be true for all such specializations of numeric_limits.

3 Non-scalar  types,  such as complex<T> (_lib.complex_), shall not have
  specializations.

  Header <limits> synopsis
  _________________________
  1) Possible definitions include 0 and 0L, but not (void*)0.

  namespace std {
    template<class T> class numeric_limits;
    enum float_rounds_style;

    class numeric_limits<bool>;

    class numeric_limits<char>;
    class numeric_limits<signed char>;
    class numeric_limits<unsigned char>;
    class numeric_limits<wchar_t>;

    class numeric_limits<short>;
    class numeric_limits<int>;
    class numeric_limits<long>;
    class numeric_limits<unsigned short>;
    class numeric_limits<unsigned int>;
    class numeric_limits<unsigned long>;

    class numeric_limits<float>;
    class numeric_limits<double>;
    class numeric_limits<long double>;
  }

  18.2.1.1  Template class numeric_limits           [lib.numeric.limits]
  namespace std {
    template<class T> class numeric_limits {
    public:
      static const bool is_specialized;
      static T min();
      static T max();
      static const int  digits;
      static const int  digits10;
      static const bool is_signed;
      static const bool is_integer;
      static const bool is_exact;
      static const int  radix;
      static T epsilon();
      static T round_error();

      static const int  min_exponent;
      static const int  min_exponent10;
      static const int  max_exponent;
      static const int  max_exponent10;

      static const bool has_infinity;
      static const bool has_quiet_NaN;
      static const bool has_signaling_NaN;
      static const bool has_denorm;
      static T infinity();
      static T quiet_NaN();
      static T signaling_NaN();
      static T denorm_min();

      static const bool is_iec559;
      static const bool is_bounded;
      static const bool is_modulo;

      static const bool traps;
      static const bool tinyness_before;
      static const float_round_style round_style;
    };
  }

1 The member is_specialized makes it  possible  to  distinguish  between
  scalar  types, which have specializations, and non-scalar types, which
  do not.

2 The members radix, epsilon(), and round_error() shall have  meaningful
  values for all floating point type specializations.

3 For  types  with  has_denorm  ==  false, the member denorm_min() shall
  return the same value as the member min().

4 The default numeric_limits<T> template shall  have  all  members,  but
  with meaningless (0 or false) values.

  +-------                 BEGIN BOX 1                -------+
  Editorial proposal:

  Add  exception  specifications of throw() to each member function, per
  95-0020/N0620.
  +-------                  END BOX 1                 -------+

  18.2.1.2  numeric_limits members          [lib.numeric.limits.members]

  static T min();

1 Minimum finite value.2)
  _________________________

2 For floating types with denormalization, returns the minimum  positive
  normalized value, denorm_min().

3 Meaningful  for  all  specializations  in which is_bounded == true, or
  is_bounded == false && is_signed == false.

  static T max();

4 Maximum finite value.3)

5 Meaningful for all specializations in which is_bounded == true.

  static const int  digits;

6 Number of radix digits which can be represented without change.

7 For built-in integer types, the number of non-sign bits in the  repre­
  sentation.

8 For   floating  point  types,  the  number  of  radix  digits  in  the
  mantissa.4)

  static const int  digits10;

9 Number of base 10 digits which can be represented without change.5)

10Meaningful for all specializations in which is_bounded == true.

  static const bool is_signed;

11True if the type is signed.

12Meaningful for all specializations.

  static const bool is_integer;

  _________________________
  2) Equivalent to CHAR_MIN, SHRT_MIN, FLT_MIN, DBL_MIN, etc.
  3) Equivalent to CHAR_MAX, SHRT_MAX, FLT_MAX, DBL_MAX, etc.
  4) Equivalent to FLT_MANT_DIG, DBL_MANT_DIG, LDBL_MANT_DIG.
  5) Equivalent to FLT_DIG, DBL_DIG, LDBL_DIG.

13True if the type is integer.

14Meaningful for all specializations.

  static const bool is_exact;

15True if the type uses an exact representation.  All integer types  are
  exact,  but  not vice versa.  For example, rational and fixed-exponent
  representations are exact but not integer.

16Meaningful for all specializations.

  static const int  radix;

17For floating types, specifies the base or radix of the exponent repre­
  sentation (often 2).6)

18For integer types, specifies the base of the representation.7)

19Meaningful for all specializations.

  static T epsilon();

20Machine epsilon:  the difference between 1 and the least value greater
  than 1 that is representable.8)

21Meaningful only for floating point types.

  static T round_error();

22Measure of the maximum rounding error.9)

  static const int  min_exponent;

23Minimum  negative  integer  such that radix raised to that power is in
  range.10)
  _________________________
  6) Equivalent to FLT_RADIX.
  7) Distinguishes types with bases other than 2 (e.g. BCD).
  8) Equivalent to FLT_EPSILON, DBL_EPSILON, LDBL_EPSILON.
  9) This has a precise definition in the  Language  Independent  Arith­
  metic (LIA-1) standard.  Required by LIA-1.
  10) Equivalent to FLT_MIN_EXP, DBL_MIN_EXP, LDBL_MIN_EXP.

24Meaningful only for floating point types.

  static const int  min_exponent10;

25Minimum negative integer such that 10  raised  to  that  power  is  in
  range.11)

26Meaningful only for floating point types.

  static const int  max_exponent;

27Maximum  positive  integer  such that radix raised to that power is in
  range.12)

28Meaningful only for floating point types.

  static const int  max_exponent10;

29Maximum positive integer such that 10  raised  to  that  power  is  in
  range.13)

30Meaningful only for floating point types.

  static const bool has_infinity;

31True if the type has a representation for positive infinity.

32Meaningful only for floating point types.

33Shall be true for all specializations in which is_iec559 == true.

  static const bool has_quiet_NaN;

34True  if  the  type  has  a representation for a quiet (non-signaling)
  ``Not a Number.''14)

  _________________________
  11) Equivalent to FLT_MIN_10_EXP, DBL_MIN_10_EXP, LDBL_MIN_10_EXP.
  12) Equivalent to FLT_MAX_EXP, DBL_MAX_EXP, LDBL_MAX_EXP.
  13) Equivalent to FLT_MAX_10_EXP, DBL_MAX_10_EXP, LDBL_MAX_10_EXP.
  14) Required by LIA-1.

35Meaningful only for floating point types.

36Shall be true for all specializations in which is_iec559 == true.

  static const bool has_signaling_NaN;

37True if the  type  has  a  representation  for  a  signaling  ``Not  a
  Number.''15)

38Meaningful only for floating point types.

39Shall be true for all specializations in which is_iec559 == true.

  static const bool has_denorm;

40True  if the type allows denormalized values (variable number of expo­
  nent bits).16)

41Meaningful only for flotaing point types.

  static T infinity();

42Representation of positive infinity, if available.17)

43Meaningful only in specializations for  which  has_infinity  ==  true.
  Required in specializations for which is_iec559 == true.

  static T quiet_NaN();

44Representation of a quiet ``Not a Number,'' if available.18)

45Meaningful  only  in  specializations for which has_quiet_NaN == true.
  Required in specializations for which is_iec559 == true.

  static T signaling_NaN();

  _________________________
  15) Required by LIA-1.
  16) Required by LIA-1.
  17) Required by LIA-1.
  18) Required by LIA-1.

46Representation of a signaling ``Not a Number,'' if available.19)

47Meaningful only in  specializations  for  which  has_signaling_NaN  ==
  true.  Required in specializations for which is_iec559 == true.

  static T denorm_min();

48Minimum positive denormalized value.20)

49Meaningful for all floating point types.

50In  specializations for which has_denorm == false, returns the minimum
  positive normalized value.

  static const bool is_iec559;

51True if and only if the type adheres to IEC 559 standard.21)

52Meaningful only for floating point types.

  static const bool is_bounded;

53True  if the set of values representable by the type is finite.22) All
  built-in types are bounded, this member would be false  for  arbitrary
  precision types.

54Meaningful for all specializations.

  static const bool is_modulo;

55True  if  the type is modulo.23) A type is modulo if it is possible to
  add two positive numbers and have a result which  wraps  around  to  a
  third number which is less.

56Generally,  this  is false for floating types, true for unsigned inte­
  gers, and true for signed integers on most machines.

  _________________________
  19) Required by LIA-1.
  20) Required by LIA-1.
  21) International Electrotechnical Commission standard 559 is the same
  as IEEE 754.
  22) Required by LIA-1.
  23) Required by LIA-1.

57Meaningful for all specializations.

  static const bool traps;

58true if trapping is implemented for the type.24)

59Meaningful for all specializations.

  static const bool tinyness_before;

60true if tinyness is detected before rounding.25)

61Meaningful only for floating point types.

  static const float_round_style round_style;

62The rounding style for the type.26)

63Meaningful for all floating point types.  Specializations for  integer
  types shall return round_toward_zero.

  18.2.1.3  Type float_round_style                     [lib.round.style]

  namespace std {
    enum float_round_style {
      round_indeterminate       = -1,
      round_toward_zero         =  0,
      round_to_nearest          =  1,
      round_toward_infinity     =  2,
      round_toward_neg_infinity =  3
    };
  }

  18.2.1.4  numeric_limits specializations         [lib.numeric.special]

1 All  members shall be provided for all specializations.  However, many
  values are only required to be  meaningful  under  certain  conditions
  (for  example,  epsilon()  is only meaningful if is_integer is false).
  Any value which is not ``meaningful'' shall be set to 0 or false.

  _________________________
  24) Required by LIA-1.
  25) Refer to IEC 559.  Required by LIA-1.
  26) Equivalent to FLT_ROUNDS.  Required by LIA-1.

2 [Example:
  namespace std {
    class numeric_limits<float> {
    public:
      static const bool is_specialized = true;
      inline static float min() { return 1.17549435E-38F; }
      inline static float max() { return 3.40282347E+38F; }
      static const int digits   = 24;
      static const int digits10 =  6;
      static const bool is_signed  = true;
      static const bool is_integer = false;
      static const bool is_exact   = false;
      static const int radix = 2;
      inline static float epsilon()     { return 1.19209290E-07F; }
      inline static float round_error() { return 0.5F; }
      static const int min_exponent   = -125;
      static const int min_exponent10 = - 37;
      static const int max_exponent   = +128;
      static const int max_exponent10 = + 38;
      static const bool has_infinity      = true;
      static const bool has_quiet_NaN     = true;
      static const bool has_signaling_NaN = true;
      static const bool has_denorm        = false;
      inline static float infinity()      { return ...; }
      inline static float quiet_NaN()     { return ...; }
      inline static float signaling_NaN() { return ...; }
      inline static float denorm_min()    { return min(); }
      static const bool is_iec559  = true;
      static const bool is_bounded = true;
      static const bool is_modulo  = false;
      static const bool traps      = true;
      static const bool tinyness_before = true;
      static const float_round_style round_style = round_to_nearest;
    };
  }
   --end example]

  18.2.2  C Library                                       [lib.c.limits]

1 Header <climits> (Table 3):

                    Table 3--Header <climits> synopsis

  +---------------------------------------------------------------------+
  |  Type                              Name(s)                          |
  +---------------------------------------------------------------------+
  |Values:                                                              |
  |CHAR_BIT   INT_MAX    LONG_MIN     SCHAR_MIN   UCHAR_MAX   USHRT_MAX |
  |CHAR_MAX   INT_MIN    MB_LEN_MAX   SHRT_MAX    UINT_MAX              |
  |CHAR_MIN   LONG_MAX   SCHAR_MAX    SHRT_MIN    ULONG_MAX             |
  +---------------------------------------------------------------------+

2 The contents are the same as the Standard C library.

3 Header <cfloat> (Table 4):

                    Table 4--Header <cfloat> synopsis

   +-------------------------------------------------------------------+
   |     Type                             Name(s)                      |
   +-------------------------------------------------------------------+
   |Values:                                                            |
   |DBL_DIG          DBL_MIN_EXP      FLT_MIN_10_EXP   LDBL_MAX_10_EXP |
   |DBL_EPSILON      FLT_DIG          FLT_MIN_EXP      LDBL_MAX_EXP    |
   |DBL_MANT_DIG     FLT_EPSILON      FLT_RADIX        LDBL_MIN        |
   |DBL_MAX          FLT_MANT_DIG     FLT_ROUNDS       LDBL_MIN_10_EXP |
   |DBL_MAX_10_EXP   FLT_MAX          LDBL_DIG         LDBL_MIN_EXP    |
   |DBL_MAX_EXP      FLT_MAX_10_EXP   LDBL_EPSILON                     |
   |DBL_MIN          FLT_MAX_EXP      LDBL_MANT_DIG                    |
   |DBL_MIN_10_EXP   FLT_MIN          LDBL_MAX                         |
   +-------------------------------------------------------------------+

4 The contents are the same as the Standard C library.

  SEE ALSO: ISO C subclause 7.1.5, 5.2.4.2.2, 5.2.4.2.1.

  18.3  Start and termination                   [lib.support.start.term]

1 Header <cstdlib> (partial), Table 5:

                    Table 5--Header <cstdlib> synopsis

               +-------------------------------------------+
               |   Type                 Name(s)            |
               +-------------------------------------------+
               |Macros:      EXIT_FAILURE     EXIT_SUCCESS |
               +-------------------------------------------+
               |Functions:   abort   atexit   exit         |
               +-------------------------------------------+

2 The contents are the same as the Standard C library, with the  follow­
  ing changes:

  atexit(void (*f)(void))

3 The  function  atexit(), has additional behavior in this International
  Standard:

  --For the execution of a function registered with atexit,  if  control
    leaves  the  function  because  it  provides no handler for a thrown

    exception, terminate() is called (_lib.terminate_).

  exit(int status)

4 The function exit() has  additional  behavior  in  this  International
  Standard:

  --First,  all  functions f registered by calling atexit(f) are called,
    in the reverse order of their registration.27)

  --Next, all static objects are destroyed in the reverse order of their
    construction.  (Automatic objects are not destroyed as a  result  of
    calling exit().)28)

  --Next, all open C streams (as mediated  by  the  function  signatures
    declared  in <cstdio>) with unwritten buffered data are flushed, all
    open C streams are closed, and all files  created  by  calling  tmp­
    file() are removed.29)

  --Finally, control is returned to the host environment.  If status  is
    zero  or  EXIT_SUCCESS, an implementation-defined form of the status
    successful termination is returned.  If status is  EXIT_FAILURE,  an
    implementation-defined  form  of the status unsuccessful termination
    is    returned.     Otherwise     the     status     returned     is
    implementation-defined.30)

5 The function exit() never returns to its caller.

  SEE ALSO: subclauses _basic.start_,  _basic.start.term_,  ISO  C  sub­
  clause 7.10.4.

  18.4  Dynamic memory management                  [lib.support.dynamic]

1 The  header <new> defines several functions that manage the allocation
  of dynamic storage in a  program.   It  also  defines  components  for
  reporting storage management errors.

  Header <new> synopsis
  _________________________
  27) A function is called for every time it is registered.   The  func­
  tion signature atexit(void (*)()), is declared in <cstdlib>.
  28)  Automatic  objects  are all destroyed in a program whose function
  main() contains no automatic objects and executes the call to  exit().
  Control  can  be  transferred directly to such a main() by throwing an
  exception that is caught in main().
  29)    Any    C    streams    associated    with    cin,   cout,   etc
  (_lib.iostream.objects_) are flushed and closed  when  static  objects
  are  destroyed  in  the previous phase.  The function tmpfile() is de­
  clared in <cstdio>.
  30) The macros EXIT_FAILURE and EXIT_SUCCESS are defined in <cstdlib>.

  #include <cstdlib>      // for size_t
  #include <stdexcept>    // for exception

  namespace std {
    void* operator new(size_t size) throw(bad_alloc);
    struct nothrow {};
    void* operator new(size_t size, const nothrow&) throw();
    void  operator delete(void* ptr) throw();
    void* operator new[](size_t size) throw(bad_alloc);
    void* operator new[](size_t size, const nothrow&) throw();
    void  operator delete[](void* ptr) throw();
    void* operator new  (size_t size, void* ptr) throw();
    void* operator new[](size_t size, void* ptr) throw();
    void  operator delete  (void* ptr, void*) throw();
    void  operator delete[](void* ptr, void*) throw();
    class bad_alloc;
    typedef void (*new_handler)();
    new_handler set_new_handler(new_handler new_p);
  }

  SEE ALSO:  subclauses _intro.memory_, _basic.stc.dynamic_, _expr.new_,
  _expr.delete_, _class.free_, subclause _lib.memory_, Memory.

  18.4.1  Storage allocation and deallocation           [lib.new.delete]

  18.4.1.1  Single-object forms                  [lib.new.delete.single]

  void* operator new(size_t size) throw(bad_alloc);

  Effects:
    The allocation function (_basic.stc.dynamic.allocation_) called by a
    new-expression  (_expr.new_) to allocate size bytes of storage suit­
    ably aligned to represent any object of that size.
  Replaceable:
    a C++ program may define a function  with  this  function  signature
    that  displaces  the  default  version  defined  by the C++ Standard
    library.
  Required behavior:
    Return    a    pointer    to    dynamically    allocated     storage
    (_basic.stc.dynamic_), or else throw a bad_alloc exception.
  Default behavior:

  --Executes  a  loop:  Within  the loop, the function first attempts to
    allocate the requested storage.  Whether the attempt involves a call
    to the Standard C library function malloc is unspecified.

  --Returns  a  pointer  to the allocated storage if the attempt is suc­
    cessful.  Otherwise, if the last argument to set_new_handler() was a
    null pointer, throw bad_alloc.

  --Otherwise,    the    function    calls   the   current   new_handler
    (_lib.new.handler_).  If  the  called  function  returns,  the  loop
    repeats.

  --The  loop terminates when an attempt to allocate the requested stor­
    age is successful or when a called  new_handler  function  does  not
    return.

  void* operator new(size_t size, const nothrow&) throw();

  Effects:
    Same  as above, except that it is called by a placement version of a
    new-expression when a C++ program prefers a null pointer  result  as
    an error indication, instead of a bad_alloc exception.
  Replaceable:
    a  C++  program  may  define a function with this function signature
    that displaces the default  version  defined  by  the  C++  Standard
    library.
  Required behavior:
    Return     a    pointer    to    dynamically    allocated    storage
    (_basic.stc.dynamic_), or else return a null pointer.
  Default behavior:

  --Executes a loop: Within the loop, the  function  first  attempts  to
    allocate the requested storage.  Whether the attempt involves a call
    to the Standard C library function malloc is unspecified.

  --Returns a pointer to the allocated storage if the  attempt  is  suc­
    cessful.  Otherwise, if the last argument to set_new_handler() was a
    null pointer, return a null pointer.

  --Otherwise,   the   function   calls    the    current    new_handler
    (_lib.new.handler_).   If  the  called  function  returns,  the loop
    repeats.

  --The loop terminates when an attempt to allocate the requested  stor­
    age  is  successful  or  when a called new_handler function does not
    return.  If the called new_handler function terminates by throwing a
    bad_alloc exception, the function returns a null pointer.

1 [Example:
  T* p1 = new T;             // throws bad_alloc if it fails
  T* p2 = new(nothrow()) T;  // returns 0 if it fails
   --end example]

  void operator delete(void* ptr) throw();

  Effects:
    The  deallocation function (_basic.stc.dynamic.deallocation_) called
    by a delete-expression to render the value of ptr invalid.
  Replaceable:
    a C++ program may define a function  with  this  function  signature
    that  displaces  the  default  version  defined  by the C++ Standard
    library.

  Required behavior:
    accept a value of ptr that is null or that was returned by  an  ear­
    lier   call   to   the  default  operator  new(size_t)  or  operator
    new(size_t,const nothrow&).
  Default behavior:

  --For a null value of ptr, do nothing.

  --Any other value of ptr shall be a value returned earlier by  a  call
    to the default operator new.31) For such a non-null  value  of  ptr,
    reclaims storage allocated by the earlier call to the default opera­
    tor new.
  Notes:
    It is  unspecified  under  what  conditions  part  or  all  of  such
    reclaimed  storage is allocated by a subsequent call to operator new
    or any of calloc, malloc, or realloc, declared in <cstdlib>.

  18.4.1.2  Array forms                           [lib.new.delete.array]

  void* operator new[](size_t size) throw(bad_alloc);

  Effects:
    The allocation function (_basic.stc.dynamic.allocation_)  called  by
    the  array  form  of  a new-expression (_expr.new_) to allocate size
    bytes of storage suitably aligned to represent any array  object  of
    that size or smaller.32)
  Replaceable:
    a C++ program can define a function  with  this  function  signature
    that  displaces  the  default  version  defined  by the C++ Standard
    library.
  Required behavior:
    Same as for operator new(size_t).
  Default behavior:
    Returns operator new(size).

  void* operator new[](size_t size, const nothrow&) throw();

  Effects:
    Same as above, except that it is called by a placement version of  a
    new-expression  when  a C++ program prefers a null pointer result as
    an error indication, instead of a bad_alloc exception.
  _________________________
  31) The value must not have been invalidated by an intervening call to
  operator delete(void*) (_lib.res.on.arguments_).
  32)  It  is not the direct responsibility of operator new[](size_t) or
  operator delete[](void*) to note the repetition count or element  size
  of  the  array.  Those operations are performed elsewhere in the array
  new and delete expressions.  The array new expression,  may,  however,
  increase  the  size argument to operator new[](size_t) to obtain space
  to store supplemental information.

  Replaceable:
    a C++ program can define a function  with  this  function  signature
    that  displaces  the  default  version  defined  by the C++ Standard
    library.
  Required behavior:
    Same as for operator new(size_t,const nothrow&).
  Default behavior:
    Returns operator new(size,nothrow()).

  void operator delete[](void* ptr) throw();

  Effects:
    The deallocation function (_basic.stc.dynamic.deallocation_)  called
    by  the array form of a delete-expression to render the value of ptr
    invalid.
  Replaceable:
    a C++ program can define a function  with  this  function  signature
    that  displaces  the  default  version  defined  by the C++ Standard
    library.
  Required behavior:
    accept a value of ptr that is null or that was returned by  an  ear­
    lier call to operator new[](size_t).
  Default behavior:

  --For a null value of ptr, does nothing.

  --Any  other  value of ptr shall be a value returned earlier by a call
    to  the default operator new[](size_t).33) For such a non-null value
    of ptr, reclaims storage  allocated  by  the  earlier  call  to  the
    default operator new[](size_t) or operator new[](size_t,nothrow).

1 It  is unspecified under what conditions part or all of such reclaimed
  storage is allocated by a subsequent call to operator  new(size_t)  or
  any of calloc, malloc, or realloc, declared in <cstdlib>.

  18.4.1.3  Placement forms                   [lib.new.delete.placement]

1 These  functions  are reserved, a C++ program may not define functions
  that   displace   the   versions   in   the   Standard   C++   library
  (_lib.constraints_).

  void* operator new(size_t size, void* ptr) throw();

  Returns:
    ptr.
  Notes:
    Intentionally performs no other action.34)

  _________________________
  33) The value must not have been invalidated by an intervening call to
  operator delete[](void*) (_lib.res.on.arguments_).

2 [Example:  This  can  be  useful for constructing an object at a known
  address:
  char place[sizeof(Something)];
  Something* p = new (place) Something();
   --end example]

  void* operator new[](size_t size, void* ptr) throw();

  Returns:
    ptr.
  Notes:
    Intentionally performs no other action.

  void operator delete(void* ptr, void*) throw();

  Effects:
    Intentionally performs no action.
  Notes:
    Default function called for a placement delete expression.   Comple­
    ments default placement new.

  void operator delete[](void* ptr, void*) throw();

  Effects:
    Intentionally performs no action.
  Notes:
    Default  function  called  for  a placement array delete expression.
    Complements default placement new[].

  18.4.2  Storage allocation errors                   [lib.alloc.errors]

  18.4.2.1  Class bad_alloc                              [lib.bad.alloc]
  namespace std {
    class bad_alloc : public exception {
    public:
      bad_alloc() throw();
      bad_alloc(const bad_alloc&) throw();
      bad_alloc& operator=(const bad_alloc&) throw();
      virtual ~bad_alloc() throw();
      virtual const char* what() const throw();
    };
  }

1 The class bad_alloc defines the type of objects thrown  as  exceptions
  by the implementation to report a failure to allocate storage.

  bad_alloc() throw();

  Effects:
    Constructs an object of class bad_alloc.

      bad_alloc(const bad_alloc&) throw();
      bad_alloc& operator=(const bad_alloc&) throw();

  Effects:
    Copies an object of class bad_alloc.
  Notes:
    The  result  of  calling  what()  on the newly constructed object is
    implementation-defined.

  virtual const char* what() const throw();

  Returns:
    An implementation-defined value.

  18.4.2.2  Type new_handler                           [lib.new.handler]

  typedef void (*new_handler)();

1 The type of a handler function to be called by operator new() or oper­
  ator new[]() (_lib.new.delete_) when they cannot satisfy a request for
  addtional storage.
  Required behavior:
    A new_handler shall perform one of the following:

  --make more storage available for allocation and then return;

  --throw an exception  of  type  bad_alloc  or  a  class  derived  from
    bad_alloc;

  --call either abort() or exit();
  Default behavior:
    The implementation's default new_handler throws an exception of type
    bad_alloc.

  18.4.2.3  set_new_handler                        [lib.set.new.handler]

  new_handler set_new_handler(new_handler new_p);

  Effects:
    Establishes  the  function  designated  by  new_p  as  the   current
    new_handler.
  Returns:
    the previous new_handler.

  18.5  Type identification                           [lib.support.rtti]

1 The  header <typeinfo> defines two types associated with type informa­
  tion generated by the implementation.  It also defines two  types  for
  reporting dynamic type identification errors.

  Header <typeinfo> synopsis

  #include <stdexcept>    // for exception

  namespace std {
    class type_info;
    class bad_cast;
    class bad_typeid;
  }

  SEE ALSO: subclauses _expr.dynamic.cast_, _expr.typeid_.

  18.5.1  Class type_info                                [lib.type.info]
  namespace std {
    class type_info {
    public:
      virtual ~type_info();
      bool operator==(const type_info& rhs) const;
      bool operator!=(const type_info& rhs) const;
      bool before(const type_info& rhs) const;
      const char* name() const;
    private:
      type_info(const type_info& rhs);
      type_info& operator=(const type_info& rhs);
    };
  }

1 The class type_info describes type information generated by the imple­
  mentation.  Objects of this class effectively store  a  pointer  to  a
  name  for  the  type,  and an encoded value suitable for comparing two
  types for equality or collating order.  The names, encoding rule,  and
  collating  sequence  for  types  are  all  unspecified  and may differ
  between programs.

  bool operator==(const type_info& rhs) const;

  Effects:
    Compares the current object with rhs.
  Returns:
    true if the two values describe the same type.

  bool operator!=(const type_info& rhs) const;

  Returns:
    !(*this == rhs).

  bool before(const type_info& rhs) const;

  Effects:
    Compares the current object with rhs.
  Returns:
    true if *this precedes rhs in the implementation's collation  order.

  const char* name() const;

  Returns:
    an implementation-defined value.
  Notes:
    The    message   may   be   a   null-terminated   multibyte   string
    (_lib.multibyte.strings_), suitable for conversion and display as  a
    wstring (_lib.wstring_, _lib.locale.codecvt_)

  type_info(const type_info& rhs);
  type_info& operator=(const type_info& rhs);

  Effects:
    Copies a type_info object.
  Notes:
    Since the copy constructor and assignment operator for type_info are
    private to the class, objects of this type cannot be copied.

  18.5.2  Class bad_cast                                  [lib.bad.cast]
  namespace std {
    class bad_cast : public exception {
    public:
      bad_cast() throw();
      bad_cast(const bad_cast&) throw();
      bad_cast& operator=(const bad_cast&) throw();
      virtual ~bad_cast() throw();
      virtual const char* what() const throw();
    };
  }

1 The class bad_cast defines the type of objects thrown as exceptions by
  the  implementation to report the execution of an invalid dynamic-cast
  expression (_expr.dynamic.cast_).

  bad_cast() throw();

  Effects:
    Constructs an object of class bad_cast.

      bad_cast(const bad_cast&) throw();
      bad_cast& operator=(const bad_cast&) throw();

  Effects:
    Copies an object of class bad_cast.
  Notes:
    The result of calling what() on  the  newly  constructed  object  is
    implementation-defined.

  virtual const char* what() const throw();

  Returns:
    An implementation-defined value.
  Notes:
    The    message   may   be   a   null-terminated   multibyte   string
    (_lib.multibyte.strings_), suitable for conversion and display as  a
    wstring (_lib.wstring_, _lib.locale.codecvt_)

  18.5.3  Class bad_typeid                              [lib.bad.typeid]
  namespace std {
    class bad_typeid : public exception {
    public:
      bad_typeid() throw();
      bad_typeid(const bad_typeid&) throw();
      bad_typeid& operator=(const bad_typeid&) throw();
      virtual ~bad_typeid() throw();
      virtual const char* what() const throw();
    };
  }

1 The  class bad_typeid defines the type of objects thrown as exceptions
  by the implementation to report a null pointer in a typeid  expression
  (_expr.typeid_).

  bad_typeid() throw();

  Effects:
    Constructs an object of class bad_typeid.

      bad_typeid(const bad_typeid&) throw();
      bad_typeid& operator=(const bad_typeid&) throw();

  Effects:
    Copies an object of class bad_typeid.
  Notes:
    The  result  of  calling  what()  on the newly constructed object is
    implementation-defined.

  virtual const char* what() const throw();

  Returns:
    An implementation-defined value.

  Notes:
    The   message   may   be   a   null-terminated   multibyte    string
    (_lib.multibyte.strings_),  suitable for conversion and display as a
    wstring (_lib.wstring_, _lib.locale.codecvt_)

  18.6  Exception handling                       [lib.support.exception]

1 The header <exception> defines several types and functions related  to
  the handling of exceptions in a C++ program.

  Header <exception> synopsis

  #include <stdexcept>    // for exception

  namespace std {
    class bad_exception;
    typedef void (*unexpected_handler)();
    unexpected_handler set_unexpected(unexpected_handler f);
    void unexpected();
    typedef void (*terminate_handler)();
    terminate_handler set_terminate(terminate_handler f);
    void terminate();
  }

  SEE ALSO: subclause _except.special_.

  18.6.1  Violating exception-                [lib.exception.unexpected]
       specifications

  18.6.1.1  Class bad_exception                      [lib.bad.exception]
  namespace std {
    class bad_exception : public exception {
    public:
      bad_exception() throw();
      bad_exception(const bad_exception&) throw();
      bad_exception& operator=(const bad_exception&) throw();
      virtual ~bad_exception() throw();
      virtual const char* what() const throw();
    };
  }

1 The class bad_exception defines the type of objects thrown  as  excep­
  tions  by  the  implementation  to report a violation of an exception-
  specification (_except.unexpected_).

  bad_exception() throw();

  Effects:
    Constructs an object of class bad_exception.

      bad_exception(const bad_exception&) throw();
      bad_exception& operator=(const bad_exception&) throw();

  Effects:
    Copies an object of class bad_exception.
  Notes:
    The result of calling what() on  the  newly  constructed  object  is
    implementation-defined.

  virtual const char* what() const throw();

  Returns:
    An implementation-defined value.
  Notes:
    The    message   may   be   a   null-terminated   multibyte   string
    (_lib.multibyte.strings_), suitable for conversion and display as  a
    wstring (_lib.wstring_, _lib.locale.codecvt_)

  18.6.1.2  Type unexpected_handler             [lib.unexpected.handler]

  typedef void (*unexpected_handler)();

1 The  type  of  a  handler function to be called by unexpected() when a
  function attempts to throw an exception not listed in  its  exception-
  specification.
  Required behavior:
    an  unexpected_handler  shall either throw an exception or terminate
    execution of the program without returning to the caller.  An  unex­
    pected_handler may perform any of the following:

  --throw an exception that satisfies the exception specification;

  --throw a bad_exception exception;

  --call terminate();

  --call either abort() or exit();
  Default behavior:
    The implementation's default unexpected_handler calls terminate().

  18.6.1.3  set_unexpected                          [lib.set.unexpected]

  unexpected_handler set_unexpected(unexpected_handler f);

  Effects:
    Establishes  the  function  designated  by  f  as  the current unex­
    pected_handler.
  Requires:
    f shall not be a null pointer.

  Returns:
    The previous unexpected_handler.

  18.6.1.4  unexpected                                  [lib.unexpected]

  void unexpected();

1 Called by the  implementation  when  a  function  with  an  exception-
  specification throws an exception that is not listed in the exception-
  specification (_except.unexpected_).
  Effects:
    Calls    the    current    unexpected_handler    handler    function
    (_lib.unexpected.handler_).

  18.6.2  Abnormal termination                 [lib.exception.terminate]

  18.6.2.1  Type terminate_handler               [lib.terminate.handler]

  typedef void (*terminate_handler)();

1 The type of a handler function to be called by terminate() when termi­
  nating exception processing.
  Required behavior:
    A terminate_handler shall terminate execution of the program without
    returning to the caller.
  Default behavior:
    The implementation's default terminate_handler calls abort().

  18.6.2.2  set_terminate                            [lib.set.terminate]

  terminate_handler set_terminate(terminate_handler f);

  Effects:
    Establishes  the  function  designated  by  f as the current handler
    function for terminating exception processing.
  Requires:
    f shall not be a null pointer.
  Returns:
    The previous terminate_handler.

  18.6.2.3  terminate                                    [lib.terminate]

  void terminate();

1 Called by the implementation when exception handling must be abandoned
  for any of several reasons (_except.terminate_).

  Effects:
    Calls     the    current    terminate_handler    handler    function
    (_lib.terminate.handler_).

  18.7  Other runtime support                      [lib.support.runtime]

1 Headers <cstdarg> (variable arguments),  <csetjmp>  (nonlocal  jumps),
  <ctime>  (system  clock clock(), time()), <csignal> (signal handling),
  and <cstdlib> (runtime environment getenv(), system()).

                    Table 5--Header <cstdarg> synopsis

                 +--------------------------------------+
                 | Type               Name(s)           |
                 +--------------------------------------+
                 |Macros:   va_arg    va_end   va_start |
                 +--------------------------------------+
                 |Type:     va_list                     |
                 +--------------------------------------+

                    Table 5--Header <csetjmp> synopsis

                          +--------------------+
                          |  Type      Name(s) |
                          +--------------------+
                          |Macro:      setjmp  |
                          +--------------------+
                          |Type:       jmp_buf |
                          +--------------------+
                          |Function:   longjmp |
                          +--------------------+

                     Table 5--Header <ctime> synopsis

                      +----------------------------+
                      |   Type         Name(s)     |
                      +----------------------------+
                      |Macros:      CLOCKS_PER_SEC |
                      +----------------------------+
                      |Types:       clock_t        |
                      +----------------------------+
                      |Functions:   clock          |
                      +----------------------------+

                    Table 5--Header <csignal> synopsis

         +-------------------------------------------------------+
         |   Type                       Name(s)                  |
         +-------------------------------------------------------+
         |Macros:      SIGABRT        SIGILL   SIGSEGV   SIG_DFL |
         |SIG_IGN      SIGFPE         SIGINT   SIGTERM   SIG_ERR |
         +-------------------------------------------------------+
         |Type:        sig_atomic_t                              |
         +-------------------------------------------------------+
         |Functions:   raise          signal                     |
         +-------------------------------------------------------+

                    Table 5--Header <cstdlib> synopsis

                      +-----------------------------+
                      |   Type          Name(s)     |
                      +-----------------------------+
                      |Functions:   getenv   system |
                      +-----------------------------+

2 The contents are the same as the Standard C library, with the  follow­
  ing changes:

3 The  function  signature  longjmp(jmp_buf  jbuf,  int  val)  has  more
  restricted behavior in this International Standard.  If any  automatic
  objects  would be destroyed by a thrown exception transferring control
  to another  (destination)  point  in  the  program,  then  a  call  to
  longjmp(jbuf,  val)  at  the throw point that transfers control to the
  same (destination) point has undefined behavior.

  SEE ALSO: ISO C subclause 7.10.4, 7.8, 7.6, 7.12.