path: root/acinclude.m4

dnl  MPFR specific autoconf macros

dnl  Copyright 2000, 2002-2018 Free Software Foundation, Inc.
dnl  Contributed by the AriC and Caramba projects, INRIA.
dnl
dnl  This file is part of the GNU MPFR Library.
dnl
dnl  The GNU MPFR Library is free software; you can redistribute it and/or modify
dnl  it under the terms of the GNU Lesser General Public License as published
dnl  by the Free Software Foundation; either version 3 of the License, or (at
dnl  your option) any later version.
dnl
dnl  The GNU MPFR Library is distributed in the hope that it will be useful, but
dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
dnl  License for more details.
dnl
dnl  You should have received a copy of the GNU Lesser General Public License
dnl  along with the GNU MPFR Library; see the file COPYING.LESSER.  If not, see
dnl  http://www.gnu.org/licenses/ or write to the Free Software Foundation, Inc.,
dnl  51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA.

dnl  autoconf 2.60 is necessary because of the use of AC_PROG_SED.
dnl  The following line allows the autoconf wrapper (when installed)
dnl  to work as expected.
dnl  If you change the required version, please update README.dev too!
AC_PREREQ(2.60)

dnl ------------------------------------------------------------
dnl You must put in MPFR_CONFIGS everything which configure MPFR
dnl except:
dnl   - Everything dealing with CC and CFLAGS in particular the ABI
dnl     but the IEEE-754 specific flags must be set here.
dnl   - Tests that depend on gmp.h (see MPFR_CHECK_DBL2INT_BUG as an example:
dnl     a function needs to be defined and called in configure.ac).
dnl   - GMP's linkage.
dnl   - Libtool stuff.
dnl   - Handling of special arguments of MPFR's configure.
AC_DEFUN([MPFR_CONFIGS],
[
AC_REQUIRE([AC_OBJEXT])
AC_REQUIRE([MPFR_CHECK_LIBM])
AC_REQUIRE([MPFR_CHECK_LIBQUADMATH])
AC_REQUIRE([AC_HEADER_TIME])
AC_REQUIRE([AC_CANONICAL_HOST])

AC_CHECK_HEADER([limits.h],, AC_MSG_ERROR([limits.h not found]))
AC_CHECK_HEADER([float.h],,  AC_MSG_ERROR([float.h not found]))
AC_CHECK_HEADER([string.h],, AC_MSG_ERROR([string.h not found]))

dnl Check for locales
AC_CHECK_HEADERS([locale.h])

dnl Check for wide characters (wchar_t and wint_t)
AC_CHECK_HEADERS([wchar.h])

dnl Check for stdargs
AC_CHECK_HEADER([stdarg.h],[AC_DEFINE([HAVE_STDARG],1,[Define if stdarg])],
  [AC_CHECK_HEADER([varargs.h],,
    AC_MSG_ERROR([stdarg.h or varargs.h not found]))])

dnl sys/fpu.h - MIPS specific
AC_CHECK_HEADERS([sys/time.h sys/fpu.h])

dnl Android has a <locale.h>, but not the following members.
AC_CHECK_MEMBERS([struct lconv.decimal_point, struct lconv.thousands_sep],,,
  [#include <locale.h>])

dnl Check how to get `alloca'
AC_FUNC_ALLOCA

dnl Define uintptr_t if not available.
AC_TYPE_UINTPTR_T

dnl va_copy macro
AC_MSG_CHECKING([how to copy va_list])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <stdarg.h>
]], [[
   va_list ap1, ap2;
   va_copy(ap1, ap2);
]])], [
   AC_MSG_RESULT([va_copy])
   AC_DEFINE(HAVE_VA_COPY)
], [AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <stdarg.h>
]], [[
   va_list ap1, ap2;
   __va_copy(ap1, ap2);
]])], [AC_DEFINE([HAVE___VA_COPY]) AC_MSG_RESULT([__va_copy])],
   [AC_MSG_RESULT([memcpy])])])

dnl FIXME: The functions memmove, memset and strtol are really needed by
dnl MPFR, but if they are implemented as macros, this is also OK (in our
dnl case).  So, we do not return an error, but their tests are currently
dnl useless.
dnl Moreover, for memmove and memset, when gcc -Werror is used, these
dnl functions are considered to be missing because of a "conflicting
dnl types for built-in function" error. Possible workarounds if the
dnl results of this test are used (thus one doesn't want an error):
dnl   * If "$GCC" is set, disable built-ins by adding -fno-builtin
dnl     to $CFLAGS for this test (this would yield a failure if such
dnl     functions are defined only as built-ins, but does this occur
dnl     in practice?).
dnl   * Enable -Werror only for the main compilation (and possibly
dnl     some particular tests) via a new configure option.
dnl gettimeofday is not defined for MinGW
AC_CHECK_FUNCS([memmove memset setlocale strtol gettimeofday signal])

dnl We cannot use AC_CHECK_FUNCS on sigaction, because while this
dnl function may be provided by the C library, its prototype and
dnl associated structure may not be available, e.g. when compiling
dnl with "gcc -std=c99".
AC_MSG_CHECKING(for sigaction and its associated structure)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <signal.h>
static int f (int (*func)(int, const struct sigaction *, struct sigaction *))
{ return 0; }
]], [[
 return f(sigaction);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_SIGACTION, 1,
    [Define if you have a working sigaction function.])
],[AC_MSG_RESULT(no)])

dnl check for long long
AC_CHECK_TYPE([long long int],
   AC_DEFINE(HAVE_LONG_LONG, 1, [Define if compiler supports long long]),,)

dnl intmax_t is C99
AC_CHECK_TYPES([intmax_t])
if test "$ac_cv_type_intmax_t" = yes; then
  AC_CACHE_CHECK([for working INTMAX_MAX], mpfr_cv_have_intmax_max, [
    saved_CPPFLAGS="$CPPFLAGS"
    CPPFLAGS="$CPPFLAGS -I$srcdir/src"
    AC_COMPILE_IFELSE([AC_LANG_PROGRAM(
        [[#include "mpfr-intmax.h"]],
        [[intmax_t x = INTMAX_MAX; (void) x;]]
      )],
      mpfr_cv_have_intmax_max=yes, mpfr_cv_have_intmax_max=no)
    CPPFLAGS="$saved_CPPFLAGS"
  ])
  if test "$mpfr_cv_have_intmax_max" = "yes"; then
    AC_DEFINE(MPFR_HAVE_INTMAX_MAX,1,[Define if you have a working INTMAX_MAX.])
  fi
  AC_CACHE_CHECK([for working printf length modifier for intmax_t],
                 mpfr_cv_printf_maxlm, [
    saved_CPPFLAGS="$CPPFLAGS"
    CPPFLAGS="$CPPFLAGS -I$srcdir/src"
    for modifier in j ll l
    do
      AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <stdio.h>
#include <string.h>
#include "mpfr-intmax.h"
]],[[
  char s[64];
  sprintf (s, "%${modifier}d %${modifier}u",
           (intmax_t) -17, (uintmax_t) 42);
  return strcmp (s, "-17 42") != 0;
]])],
       mpfr_cv_printf_maxlm=${modifier}; break,
       mpfr_cv_printf_maxlm=none,
dnl We assume that j is working when cross-compiling.
       mpfr_cv_printf_maxlm=j; break
      )
    done
    CPPFLAGS="$saved_CPPFLAGS"
  ])
  if test "$mpfr_cv_printf_maxlm" != "none"; then
    AC_DEFINE_UNQUOTED([MPFR_PRINTF_MAXLM],["$mpfr_cv_printf_maxlm"],
      [Define to a working printf length modifier for intmax_t])
  fi
fi

AC_CHECK_TYPE( [union fpc_csr],
   AC_DEFINE(HAVE_FPC_CSR,1,[Define if union fpc_csr is available]), ,
[
#if HAVE_SYS_FPU_H
#  include <sys/fpu.h>
#endif
])

dnl Check for _Noreturn function specifier (ISO C11)
AC_CACHE_CHECK([for _Noreturn], mpfr_cv_have_noreturn, [
  AC_COMPILE_IFELSE([AC_LANG_SOURCE([[_Noreturn void foo(int);]])],
    mpfr_cv_have_noreturn=yes, mpfr_cv_have_noreturn=no)
])
if test "$mpfr_cv_have_noreturn" = "yes"; then
  AC_DEFINE(MPFR_HAVE_NORETURN,1,[Define if the _Noreturn function specifier is supported.])
fi

dnl Check for __builtin_unreachable
AC_CACHE_CHECK([for __builtin_unreachable], mpfr_cv_have_builtin_unreachable,
[
  AC_LINK_IFELSE([AC_LANG_PROGRAM(
      [[int x;]],
      [[if (x) __builtin_unreachable(); ]]
    )],
    mpfr_cv_have_builtin_unreachable=yes,
    mpfr_cv_have_builtin_unreachable=no)
])
if test "$mpfr_cv_have_builtin_unreachable" = "yes"; then
  AC_DEFINE(MPFR_HAVE_BUILTIN_UNREACHABLE, 1,
   [Define if the __builtin_unreachable GCC built-in is supported.])
fi

dnl Check for attribute constructor and destructor
MPFR_CHECK_CONSTRUCTOR_ATTR()

dnl Check for POSIX Thread. Since the AX_PTHREAD macro is not standard
dnl (it is provided by autoconf-archive), we need to detect whether it
dnl is left unexpanded, otherwise the configure script won't fail and
dnl "make distcheck" won't give any error, yielding buggy tarballs!
dnl The \b is necessary to avoid an error with recent ax_pthread.m4
dnl (such as with Debian's autoconf-archive 20160320-1), which contains
dnl AX_PTHREAD_ZOS_MISSING, etc. It is not documented, but see:
dnl   https://lists.gnu.org/archive/html/autoconf/2015-03/msg00011.html
dnl
dnl Note: each time a change is done in m4_pattern_forbid, autogen.sh
dnl should be tested with and without ax_pthread.m4 availability (in
dnl the latter case, there should be an error).
m4_pattern_forbid([AX_PTHREAD\b])
AX_PTHREAD([])

dnl Check for ISO C11 Thread
MPFR_CHECK_C11_THREAD()

dnl Check for fesetround
AC_CACHE_CHECK([for fesetround], mpfr_cv_have_fesetround, [
saved_LIBS="$LIBS"
LIBS="$LIBS $MPFR_LIBM"
AC_LINK_IFELSE([AC_LANG_PROGRAM([[#include <fenv.h>]], [[fesetround(FE_TONEAREST);]])],
  mpfr_cv_have_fesetround=yes, mpfr_cv_have_fesetround=no)
LIBS="$saved_LIBS"
])
if test "$mpfr_cv_have_fesetround" = "yes"; then
  AC_DEFINE(MPFR_HAVE_FESETROUND,1,[Define if you have the `fesetround' function via the <fenv.h> header file.])
fi

dnl Check for gcc float-conversion bug; if need be, -ffloat-store is used to
dnl force the conversion to the destination type when a value is stored to
dnl a variable (see ISO C99 standard 5.1.2.3#13, 6.3.1.5#2, 6.3.1.8#2). This
dnl is important concerning the exponent range. Note that this doesn't solve
dnl the double-rounding problem.
if test -n "$GCC"; then
  AC_CACHE_CHECK([for gcc float-conversion bug], mpfr_cv_gcc_floatconv_bug, [
  saved_LIBS="$LIBS"
  LIBS="$LIBS $MPFR_LIBM"
  AC_RUN_IFELSE([AC_LANG_SOURCE([[
#include <float.h>
#ifdef MPFR_HAVE_FESETROUND
#include <fenv.h>
#endif
static double get_max (void);
int main (void) {
  double x = 0.5;
  double y;
  int i;
  for (i = 1; i <= 11; i++)
    x *= x;
  if (x != 0)
    return 1;
#ifdef MPFR_HAVE_FESETROUND
  /* Useful test for the G4 PowerPC */
  fesetround(FE_TOWARDZERO);
  x = y = get_max ();
  x *= 2.0;
  if (x != y)
    return 1;
#endif
  return 0;
}
static double get_max (void) { static volatile double d = DBL_MAX; return d; }
  ]])],
     [mpfr_cv_gcc_floatconv_bug="no"],
     [mpfr_cv_gcc_floatconv_bug="yes, use -ffloat-store"],
     [mpfr_cv_gcc_floatconv_bug="cannot test, use -ffloat-store"])
  LIBS="$saved_LIBS"
  ])
  if test "$mpfr_cv_gcc_floatconv_bug" != "no"; then
    CFLAGS="$CFLAGS -ffloat-store"
  fi
fi

dnl Check if subnormal (denormalized) numbers are supported
dnl for the binary64 format, the smallest normal number is 2^(-1022)
dnl for the binary32 format, the smallest normal number is 2^(-126)
AC_CACHE_CHECK([for subnormal double-precision numbers],
mpfr_cv_have_denorms, [
AC_RUN_IFELSE([AC_LANG_SOURCE([[
#include <stdio.h>
int main (void) {
  double x = 2.22507385850720138309e-308;
  fprintf (stderr, "%e\n", x / 2.0);
  return 2.0 * (double) (x / 2.0) != x;
}
]])],
   [mpfr_cv_have_denorms="yes"],
   [mpfr_cv_have_denorms="no"],
   [mpfr_cv_have_denorms="cannot test, assume no"])
])
if test "$mpfr_cv_have_denorms" = "yes"; then
  AC_DEFINE(HAVE_DENORMS,1,[Define if subnormal (denormalized) doubles work.])
fi
AC_CACHE_CHECK([for subnormal single-precision numbers],
mpfr_cv_have_denorms_flt, [
AC_RUN_IFELSE([AC_LANG_SOURCE([[
#include <stdio.h>
int main (void) {
  float x = 1.17549435082229e-38;
  fprintf (stderr, "%e\n", x / 2.0);
  return 2.0 * (float) (x / 2.0) != x;
}
]])],
   [mpfr_cv_have_denorms_flt="yes"],
   [mpfr_cv_have_denorms_flt="no"],
   [mpfr_cv_have_denorms_flt="cannot test, assume no"])
])
if test "$mpfr_cv_have_denorms_flt" = "yes"; then
  AC_DEFINE(HAVE_DENORMS_FLT,1,
  [Define if subnormal (denormalized) floats work.])
fi

dnl Check if signed zeros are supported. Note: the test will fail
dnl if the division by 0 generates a trap.
AC_CACHE_CHECK([for signed zeros], mpfr_cv_have_signedz, [
AC_RUN_IFELSE([AC_LANG_SOURCE([[
int main (void) {
  return 1.0 / 0.0 == 1.0 / -0.0;
}
]])],
   [mpfr_cv_have_signedz="yes"],
   [mpfr_cv_have_signedz="no"],
   [mpfr_cv_have_signedz="cannot test, assume no"])
])
if test "$mpfr_cv_have_signedz" = "yes"; then
  AC_DEFINE(HAVE_SIGNEDZ,1,[Define if signed zeros are supported.])
fi

dnl Check the FP division by 0 fails (e.g. on a non-IEEE-754 platform).
dnl In such a case, MPFR_ERRDIVZERO is defined to disable the tests
dnl involving a FP division by 0.
dnl For the developers: to check whether all these tests are disabled,
dnl configure MPFR with "-DMPFR_TESTS_FPE_DIV -DMPFR_ERRDIVZERO".
AC_CACHE_CHECK([if the FP division by 0 fails], mpfr_cv_errdivzero, [
AC_RUN_IFELSE([AC_LANG_SOURCE([[
int main (void) {
  volatile double d = 0.0, x;
  x = 0.0 / d;
  x = 1.0 / d;
  (void) x;
  return 0;
}
]])],
   [mpfr_cv_errdivzero="no"],
   [mpfr_cv_errdivzero="yes"],
   [mpfr_cv_errdivzero="cannot test, assume no"])
])
if test "$mpfr_cv_errdivzero" = "yes"; then
  AC_DEFINE(MPFR_ERRDIVZERO,1,[Define if the FP division by 0 fails.])
  AC_MSG_WARN([The floating-point division by 0 fails instead of])
  AC_MSG_WARN([returning a special value: NaN or infinity. Tests])
  AC_MSG_WARN([involving a FP division by 0 will be disabled.])
fi

dnl Check whether NAN != NAN (as required by the IEEE-754 standard,
dnl but not by the ISO C standard). For instance, this is false with
dnl MIPSpro 7.3.1.3m under IRIX64. By default, assume this is true.
AC_CACHE_CHECK([if NAN == NAN], mpfr_cv_nanisnan, [
AC_RUN_IFELSE([AC_LANG_SOURCE([[
#include <stdio.h>
#include <math.h>
#ifndef NAN
# define NAN (0.0/0.0)
#endif
int main (void) {
  double d;
  d = NAN;
  return d != d;
}
]])],
   [mpfr_cv_nanisnan="yes"],
   [mpfr_cv_nanisnan="no"],
   [mpfr_cv_nanisnan="cannot test, assume no"])
])
if test "$mpfr_cv_nanisnan" = "yes"; then
  AC_DEFINE(MPFR_NANISNAN,1,[Define if NAN == NAN.])
  AC_MSG_WARN([The test NAN != NAN is false. The probable reason is that])
  AC_MSG_WARN([your compiler optimizes floating-point expressions in an])
  AC_MSG_WARN([unsafe way because some option, such as -ffast-math or])
  AC_MSG_WARN([-fast (depending on the compiler), has been used.  You])
  AC_MSG_WARN([should NOT use such an option, otherwise MPFR functions])
  AC_MSG_WARN([such as mpfr_get_d and mpfr_set_d may return incorrect])
  AC_MSG_WARN([results on special FP numbers (e.g. NaN or signed zeros).])
  AC_MSG_WARN([If you did not use such an option, please send us a bug])
  AC_MSG_WARN([report so that we can try to find a workaround for your])
  AC_MSG_WARN([platform and/or document the behavior.])
fi

dnl Check if the chars '0' to '9' are consecutive values
AC_MSG_CHECKING([if charset has consecutive values])
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
char *number = "0123456789";
char *lower  = "abcdefghijklmnopqrstuvwxyz";
char *upper  = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
]],[[
 int i;
 unsigned char *p;
 for (p = (unsigned char*) number, i = 0; i < 9; i++)
   if ( (*p)+1 != *(p+1) ) return 1;
 for (p = (unsigned char*) lower, i = 0; i < 25; i++)
   if ( (*p)+1 != *(p+1) ) return 1;
 for (p = (unsigned char*) upper, i = 0; i < 25; i++)
   if ( (*p)+1 != *(p+1) ) return 1;
]])], [AC_MSG_RESULT(yes)],[
 AC_MSG_RESULT(no)
 AC_DEFINE(MPFR_NO_CONSECUTIVE_CHARSET,1,[Charset is not consecutive])
], [AC_MSG_RESULT(cannot test)])

dnl Must be checked with the LIBM
dnl but we don't want to add the LIBM to MPFR dependency.
dnl Can't use AC_CHECK_FUNCS since the function may be in LIBM but
dnl not exported in math.h
saved_LIBS="$LIBS"
LIBS="$LIBS $MPFR_LIBM"
dnl AC_CHECK_FUNCS([round trunc floor ceil nearbyint])
AC_MSG_CHECKING(for math/round)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <math.h>
static int f (double (*func)(double)) { return 0; }
]], [[
 return f(round);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_ROUND, 1,[Have ISO C99 round function])
],[AC_MSG_RESULT(no)])

AC_MSG_CHECKING(for math/trunc)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <math.h>
static int f (double (*func)(double)) { return 0; }
]], [[
 return f(trunc);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_TRUNC, 1,[Have ISO C99 trunc function])
],[AC_MSG_RESULT(no)])

AC_MSG_CHECKING(for math/floor)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <math.h>
static int f (double (*func)(double)) { return 0; }
]], [[
 return f(floor);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_FLOOR, 1,[Have ISO C99 floor function])
],[AC_MSG_RESULT(no)])

AC_MSG_CHECKING(for math/ceil)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <math.h>
static int f (double (*func)(double)) { return 0; }
]], [[
 return f(ceil);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_CEIL, 1,[Have ISO C99 ceil function])
],[AC_MSG_RESULT(no)])

AC_MSG_CHECKING(for math/nearbyint)
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <math.h>
static int f (double (*func)(double)) { return 0; }
]], [[
 return f(nearbyint);
]])], [
   AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_NEARBYINT, 1,[Have ISO C99 nearbyint function])
],[AC_MSG_RESULT(no)])

dnl Check if _mulx_u64 is provided
dnl Note: This intrinsic is not standard. We need a run because
dnl it may be provided but not working as expected (with ICC 15,
dnl one gets an "Illegal instruction").
AC_MSG_CHECKING([for _mulx_u64])
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <immintrin.h>
]], [[
 unsigned long long h1, h2;
 _mulx_u64(17, 42, &h1);
 _mulx_u64(-1, -1, &h2);
 return h1 == 0 && h2 == -2 ? 0 : 1;
]])],
  [AC_MSG_RESULT(yes)
   AC_DEFINE(HAVE_MULX_U64, 1,[Have a working _mulx_u64 function])
  ],
  [AC_MSG_RESULT(no)
  ],
  [AC_MSG_RESULT([cannot test, assume no])
  ])

LIBS="$saved_LIBS"

dnl Now try to check the long double format
MPFR_C_LONG_DOUBLE_FORMAT

if test "$enable_logging" = yes; then
  if test "$enable_thread_safe" = yes; then
    AC_MSG_ERROR([enable either `Logging' or `thread-safe', not both])
  else
    enable_thread_safe=no
  fi
fi

dnl Check if thread-local variables are supported.
dnl At least two problems can occur in practice:
dnl 1. The compilation fails, e.g. because the compiler doesn't know
dnl    about the __thread keyword.
dnl 2. The compilation succeeds, but the system doesn't support TLS or
dnl    there is some ld configuration problem. One of the effects can
dnl    be that thread-local variables always evaluate to 0. So, it is
dnl    important to run the test below.
if test "$enable_thread_safe" != no; then
AC_MSG_CHECKING(for TLS support using C11)
saved_CPPFLAGS="$CPPFLAGS"
CPPFLAGS="$CPPFLAGS -I$srcdir/src"
AC_RUN_IFELSE([AC_LANG_SOURCE([[
#define MPFR_USE_THREAD_SAFE 1
#define MPFR_USE_C11_THREAD_SAFE 1
#include "mpfr-thread.h"
MPFR_THREAD_ATTR int x = 17;
int main (void) {
  return x != 17;
}
  ]])],
     [AC_MSG_RESULT(yes)
      AC_DEFINE([MPFR_USE_THREAD_SAFE],1,[Build MPFR as thread safe])
      AC_DEFINE([MPFR_USE_C11_THREAD_SAFE],1,[Build MPFR as thread safe using C11])
      tls_c11_support=yes
      enable_thread_safe=yes
     ],
     [AC_MSG_RESULT(no)
     ],
     [AC_MSG_RESULT([cannot test, assume no])
     ])
CPPFLAGS="$saved_CPPFLAGS"

if test "$tls_c11_support" != "yes"
then

 AC_MSG_CHECKING(for TLS support)
 saved_CPPFLAGS="$CPPFLAGS"
 CPPFLAGS="$CPPFLAGS -I$srcdir/src"
 AC_RUN_IFELSE([AC_LANG_SOURCE([[
 #define MPFR_USE_THREAD_SAFE 1
 #include "mpfr-thread.h"
 MPFR_THREAD_ATTR int x = 17;
 int main (void) {
   return x != 17;
 }
   ]])],
      [AC_MSG_RESULT(yes)
       AC_DEFINE([MPFR_USE_THREAD_SAFE],1,[Build MPFR as thread safe])
       enable_thread_safe=yes
      ],
      [AC_MSG_RESULT(no)
       if test "$enable_thread_safe" = yes; then
         AC_MSG_ERROR([please configure with --disable-thread-safe])
       fi
      ],
      [if test "$enable_thread_safe" = yes; then
         AC_MSG_RESULT([cannot test, assume yes])
         AC_DEFINE([MPFR_USE_THREAD_SAFE],1,[Build MPFR as thread safe])
       else
         AC_MSG_RESULT([cannot test, assume no])
       fi
      ])
 CPPFLAGS="$saved_CPPFLAGS"
 fi
fi

dnl Check if decimal floats are available.
dnl For the different cases, we try to use values that will not be returned
dnl by build tools. For instance, 1 must not be used as it can be returned
dnl by ld in case of link failure.
dnl Note: We currently reread the 64-bit data memory as a double and compare
dnl it with constants. However, if there is any issue with double, such as
dnl the use of an extended precision, this may fail. Possible solutions:
dnl   1. Use the hex format for the double constants (this format should be
dnl      supported if _Decimal64 is).
dnl   2. Use more precision in the double constants (more decimal digits),
dnl      just in case.
dnl   3. Use uint64_t (or unsigned long long, though this type might not be
dnl      on 64 bits) instead of or in addition to the test on double.
dnl   4. Use an array of 8 unsigned char's instead of or in addition to the
dnl      test on double, considering the 2 practical cases of endianness.
if test "$enable_decimal_float" != no; then
  AC_MSG_CHECKING(if compiler knows _Decimal64)
  AC_COMPILE_IFELSE(
    [AC_LANG_PROGRAM([[_Decimal64 x;]])],
    [AC_MSG_RESULT(yes)
     AC_MSG_CHECKING(decimal float format)
     AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <stdlib.h>
]], [[
volatile _Decimal64 x = 1;
union { double d; _Decimal64 d64; } y;
if (x != x) return 83;
y.d64 = 1234567890123456.0dd;
return y.d == 0.14894469406741037E-123 ? 80 :
       y.d == 0.59075095508629822E-68  ? 81 : 82;
]])], [AC_MSG_RESULT(internal error)
       AC_MSG_FAILURE(unexpected exit status 0)],
      [d64_exit_status=$?
       case "$d64_exit_status" in
         80) AC_MSG_RESULT(DPD)
             AC_DEFINE([DPD_FORMAT],1,[DPD format])
             AC_DEFINE([MPFR_WANT_DECIMAL_FLOATS],1,
                       [Build decimal float functions]) ;;
         81) AC_MSG_RESULT(BID)
             AC_DEFINE([MPFR_WANT_DECIMAL_FLOATS],1,
                       [Build decimal float functions]) ;;
         82) AC_MSG_RESULT(neither DPD nor BID)
             if test "$enable_decimal_float" = yes; then
               AC_MSG_ERROR([unsupported decimal float format.
Please build MPFR without --enable-decimal-float.])
             fi ;;
         *) AC_MSG_RESULT(unknown (exit status $d64_exit_status))
             if test "$enable_decimal_float" = yes; then
               AC_MSG_ERROR([internal or link error.
Please build MPFR without --enable-decimal-float.])
             fi ;;
       esac],
      [AC_MSG_RESULT(assuming DPD)
       AC_DEFINE([DPD_FORMAT],1,[])
       AC_DEFINE([MPFR_WANT_DECIMAL_FLOATS],1,
                 [Build decimal float functions])])
              ],
    [AC_MSG_RESULT(no)
     if test "$enable_decimal_float" = yes; then
       AC_MSG_ERROR([compiler doesn't know _Decimal64 (ISO/IEC TR 24732).
Please use another compiler or build MPFR without --enable-decimal-float.])
     fi])
fi

dnl Check the bit-field ordering for _Decimal128.
dnl Little endian: sig=0 comb=49400 t0=0 t1=0 t2=0 t3=10
dnl Big endian: sig=0 comb=8 t0=0 t1=0 t2=0 t3=570933248
AC_MSG_CHECKING(bit-field ordering for _Decimal128)
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
]], [[
  union ieee_decimal128
  {
    struct
    {
      unsigned int t3:32;
      unsigned int t2:32;
      unsigned int t1:32;
      unsigned int t0:14;
      unsigned int comb:17;
      unsigned int sig:1;
    } s;
    _Decimal128 d128;
  } x;

  x.d128 = 1.0dl;
  if (x.s.sig==0 && x.s.comb==49400 && x.s.t0==0 && x.s.t1==0 && x.s.t2==0 && x.s.t3==10)
     return 1; /* little endian */
  else if (x.s.sig==0 && x.s.comb==8 && x.s.t0==0 && x.s.t1==0 && x.s.t2==0 && x.s.t3==570933248)
     return 2; /* big endian */
  else
     return 0; /* unknown encoding */
]])], [AC_MSG_RESULT(unknown)],
      [d128_exit_status=$?
       case "$d128_exit_status" in
         1) AC_MSG_RESULT(little endian)
            AC_DEFINE([HAVE_DECIMAL128_IEEE_LITTLE_ENDIAN],1) ;;
         2) AC_MSG_RESULT(big endian)
            AC_DEFINE([HAVE_DECIMAL128_IEEE_BIG_ENDIAN],1) ;;
       esac],
      [AC_MSG_RESULT(cannot test)])
  
dnl Check if _Float128 or __float128 is available. We also require the
dnl compiler to support hex constants with the f128 or q suffix (this
dnl prevents the _Float128 support with GCC's -std=c90, but who cares?).
dnl Note: We use AC_LINK_IFELSE instead of AC_COMPILE_IFELSE since an
dnl error may occur only at link time, such as under NetBSD:
dnl   https://mail-index.netbsd.org/pkgsrc-users/2018/02/02/msg026220.html
dnl   https://mail-index.netbsd.org/pkgsrc-users/2018/02/05/msg026238.html
dnl By using volatile and making the exit code depend on the value of
dnl this variable, we also make sure that optimization doesn't make
dnl the "undefined reference" error disappear.
if test "$enable_float128" != no; then
   AC_MSG_CHECKING(if _Float128 with hex constants is supported)
   AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[
volatile _Float128 x = 0x1.fp+16383f128;
return x == 0;
]])],
      [AC_MSG_RESULT(yes)
       AC_DEFINE([MPFR_WANT_FLOAT128],1,[Build float128 functions])],
      [AC_MSG_RESULT(no)
       AC_MSG_CHECKING(if __float128 can be used as a fallback)
dnl Use the q suffix in this case.
       AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#define _Float128 __float128
]], [[
volatile _Float128 x = 0x1.fp+16383q;
return x == 0;
]])],
          [AC_MSG_RESULT(yes)
           AC_DEFINE([MPFR_WANT_FLOAT128],2,
                     [Build float128 functions with float128 fallback])
           AC_DEFINE([_Float128],[__float128],[__float128 fallback])],
          [AC_MSG_RESULT(no)
           if test "$enable_float128" = yes; then
              AC_MSG_ERROR(
[compiler doesn't know _Float128 or __float128 with hex constants.
Please use another compiler or build MPFR without --enable-float128.])
       fi])
      ])
fi

dnl Check if Static Assertions are supported.
AC_MSG_CHECKING(for Static Assertion support)
saved_CPPFLAGS="$CPPFLAGS"
CPPFLAGS="$CPPFLAGS -I$srcdir/src"
AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
#define MPFR_USE_STATIC_ASSERT 1
#include "mpfr-sassert.h"

/* Test if Static Assertions work */
MPFR_DECL_STATIC_ASSERT(sizeof(char) <= sizeof(int));

int main (void) {
  MPFR_DECL_STATIC_ASSERT(sizeof(int) <= sizeof(long));
  int x;
  (void) (x = 1);  /* cast to void: avoid a warning, at least with GCC */
  /* Test of the macro after a declaraction and a statement. */
  MPFR_STAT_STATIC_ASSERT(sizeof(short) <= sizeof(int));
  return 0;
}
  ]])],
     [AC_MSG_RESULT(yes)
      AC_DEFINE([MPFR_USE_STATIC_ASSERT],1,[Build MPFR with Static Assertions])
     ],
     [AC_MSG_RESULT(no)
     ])
CPPFLAGS="$saved_CPPFLAGS"

if test "$enable_lto" = "yes" ; then
   MPFR_LTO
fi

dnl Check if the shared cache was requested and its requirements are ok.
if test "$mpfr_want_shared_cache" = yes ;then
   MPFR_CHECK_SHARED_CACHE()
fi

])
dnl end of MPFR_CONFIGS


dnl MPFR_CHECK_GMP
dnl --------------
dnl Check GMP library vs header. Useful if the user provides --with-gmp
dnl with a directory containing a GMP version that doesn't have the
dnl correct ABI: the previous tests won't trigger the error if the same
dnl GMP version with the right ABI is installed on the system, as this
dnl library is automatically selected by the linker, while the header
dnl (which depends on the ABI) of the --with-gmp include directory is
dnl used.
dnl Note: if the error is changed to a warning due to that fact that
dnl libtool is not used, then the same thing should be done for the
dnl other tests based on GMP.
AC_DEFUN([MPFR_CHECK_GMP], [
AC_REQUIRE([MPFR_CONFIGS])dnl
AC_CACHE_CHECK([for GMP library vs header correctness], mpfr_cv_check_gmp, [
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <stdio.h>
#include <limits.h>
#include <gmp.h>
]], [[
  fprintf (stderr, "GMP_NAIL_BITS     = %d\n", (int) GMP_NAIL_BITS);
  fprintf (stderr, "GMP_NUMB_BITS     = %d\n", (int) GMP_NUMB_BITS);
  fprintf (stderr, "mp_bits_per_limb  = %d\n", (int) mp_bits_per_limb);
  fprintf (stderr, "sizeof(mp_limb_t) = %d\n", (int) sizeof(mp_limb_t));
  if (GMP_NAIL_BITS != 0)
    {
      fprintf (stderr, "GMP_NAIL_BITS != 0\n");
      return 1;
    }
  if (GMP_NUMB_BITS != mp_bits_per_limb)
    {
      fprintf (stderr, "GMP_NUMB_BITS != mp_bits_per_limb\n");
      return 2;
    }
  if (GMP_NUMB_BITS != sizeof(mp_limb_t) * CHAR_BIT)
    {
      fprintf (stderr, "GMP_NUMB_BITS != sizeof(mp_limb_t) * CHAR_BIT\n");
      return 3;
    }
  return 0;
]])], [mpfr_cv_check_gmp="yes"],
      [mpfr_cv_check_gmp="no (exit status is $?)"],
      [mpfr_cv_check_gmp="cannot test, assume yes"])
])
case $mpfr_cv_check_gmp in
no*)
  AC_MSG_ERROR([bad GMP library or header - ABI problem?
See 'config.log' for details.]) ;;
esac
])


dnl MPFR_CHECK_DBL2INT_BUG
dnl ----------------------
dnl Check for double-to-integer conversion bug
dnl https://gforge.inria.fr/tracker/index.php?func=detail&aid=14435
dnl For the exit status, the lowest values (including some values after 128)
dnl are reserved for various system errors. So, let's use the largest values
dnl below 255 for errors in the test itself.
dnl The following problem has been seen under Solaris in config.log,
dnl i.e. the failure to link with libgmp wasn't detected in the first
dnl test:
dnl   configure: checking if gmp.h version and libgmp version are the same
dnl   configure: gcc -o conftest -Wall -Wmissing-prototypes [...]
dnl   configure: $? = 0
dnl   configure: ./conftest
dnl   ld.so.1: conftest: fatal: libgmp.so.10: open failed: No such file [...]
dnl   configure: $? = 0
dnl   configure: result: yes
dnl   configure: checking for double-to-integer conversion bug
dnl   configure: gcc -o conftest -Wall -Wmissing-prototypes [...]
dnl   configure: $? = 0
dnl   configure: ./conftest
dnl   ld.so.1: conftest: fatal: libgmp.so.10: open failed: No such file [...]
dnl   ./configure[1680]: eval: line 1: 1971: Killed
dnl   configure: $? = 9
dnl   configure: program exited with status 9
AC_DEFUN([MPFR_CHECK_DBL2INT_BUG], [
AC_REQUIRE([MPFR_CONFIGS])dnl
AC_CACHE_CHECK([for double-to-integer conversion bug], mpfr_cv_dbl_int_bug, [
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <gmp.h>
]], [[
  double d;
  mp_limb_t u;
  int i;

  d = 1.0;
  for (i = 0; i < GMP_NUMB_BITS - 1; i++)
    d = d + d;
  u = (mp_limb_t) d;
  for (; i > 0; i--)
    {
      if (u & 1)
        break;
      u = u >> 1;
    }
  return (i == 0 && u == 1UL) ? 0 : 254 - i;
]])], [mpfr_cv_dbl_int_bug="no"],
      [mpfr_cv_dbl_int_bug="yes or failed to exec (exit status is $?)"],
      [mpfr_cv_dbl_int_bug="cannot test, assume not present"])
])
case $mpfr_cv_dbl_int_bug in
yes*)
  AC_MSG_ERROR([double-to-integer conversion is incorrect.
You need to use another compiler (or lower the optimization level).]) ;;
esac
])

dnl MPFR_CHECK_MP_LIMB_T_VS_LONG
dnl ----------------------------
dnl Check that a long can fit in a mp_limb_t.
dnl If so, it set the define MPFR_LONG_WITHIN_LIMB
dnl FIXME: The following code is wrong when using:
dnl   -std=c99 -pedantic-errors -Wno-error=overlength-strings
dnl because static assertions are not supported, so that MPFR_ASSERTN
dnl is used, but it is not defined by "mpfr-sassert.h". The consequence
dnl is that the program fails, and MPFR_LONG_WITHIN_LIMB is not defined
dnl while it should be. Before fixing this bug, find a way to test with
dnl MPFR_LONG_WITHIN_LIMB undefined; this is necessary for code coverage
dnl and possibly to find new bugs, like in r12252:
dnl   ./configure --enable-assert=full \
dnl     'CFLAGS=-std=c99 -O3 -pedantic-errors -Wno-error=overlength-strings'
dnl According to the GMP developers, a limb is always as large as a long,
dnl except when __GMP_SHORT_LIMB is defined, but this is never defined:
dnl https://gmplib.org/list-archives/gmp-discuss/2018-February/006190.html
dnl FIXME: This is not safe. The fact that __GMP_SHORT_LIMB cannot occur
dnl is not documented, and gmp.h has such a code probably because it may
dnl occur in the future (or the user may want to override the default
dnl choice).
AC_DEFUN([MPFR_CHECK_MP_LIMB_T_VS_LONG], [
AC_REQUIRE([MPFR_CONFIGS])
AC_CACHE_CHECK([for long to fit in mp_limb_t], mpfr_cv_long_within_limb, [
saved_CPPFLAGS="$CPPFLAGS"
CPPFLAGS="$CPPFLAGS -I$srcdir/src"
AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[
#include <gmp.h>
#include "mpfr-sassert.h"
]], [[
  MPFR_STAT_STATIC_ASSERT ((mp_limb_t) -1 >= (unsigned long) -1);
  return 0;
]])], [mpfr_cv_long_within_limb="yes"],
      [mpfr_cv_long_within_limb="no"])
])
case $mpfr_cv_long_within_limb in
yes*)
      AC_DEFINE([MPFR_LONG_WITHIN_LIMB],1,[long can be stored in mp_limb_t]) ;;
esac
CPPFLAGS="$saved_CPPFLAGS"
])

dnl MPFR_CHECK_MP_LIMB_T_VS_INTMAX
dnl ------------------------------
dnl Check that an intmax_t can fit in a mp_limb_t.
AC_DEFUN([MPFR_CHECK_MP_LIMB_T_VS_INTMAX], [
AC_REQUIRE([MPFR_CONFIGS])
AC_CACHE_CHECK([for intmax_t to fit in mp_limb_t], mpfr_cv_intmax_within_limb, [
saved_CPPFLAGS="$CPPFLAGS"
CPPFLAGS="$CPPFLAGS -I$srcdir/src"
dnl AC_LINK_IFELSE is safier than AC_COMPILE_IFELSE, which did not detect
dnl the missing #include "mpfr-sassert.h".
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <gmp.h>
#include "mpfr-sassert.h"
#include "mpfr-intmax.h"
]], [[
  MPFR_STAT_STATIC_ASSERT ((mp_limb_t) -1 >= (uintmax_t) -1);
  return 0;
]])], [mpfr_cv_intmax_within_limb="yes"],
      [mpfr_cv_intmax_within_limb="no"])
])
case $mpfr_cv_intmax_within_limb in
yes*)
      AC_DEFINE([MPFR_INTMAX_WITHIN_LIMB],1,[intmax_t can be stored in mp_limb_t]) ;;
esac
CPPFLAGS="$saved_CPPFLAGS"
])

dnl MPFR_PARSE_DIRECTORY
dnl Input:  $1 = a string to a relative or absolute directory
dnl Output: $2 = the variable to set with the absolute directory
AC_DEFUN([MPFR_PARSE_DIRECTORY],
[
 dnl Check if argument is a directory
 if test -d $1 ; then
    dnl Get the absolute path of the directory in case of relative directory
    dnl with the realpath command. If the output is empty, the cause may be
    dnl that this command has not been found, and we do an alternate test,
    dnl the same as what autoconf does for the generated configure script to
    dnl determine whether a pathname is absolute or relative.
    local_tmp=`realpath $1 2>/dev/null`
    if test "$local_tmp" != "" ; then
       if test -d "$local_tmp" ; then
           $2="$local_tmp"
       else
           $2=$1
       fi
    else
       dnl The quadrigraphs @<:@, @:>@ and @:}@ produce [, ] and )
       dnl respectively (see Autoconf manual). We cannot use quoting here
       dnl as the result depends on the context in which this macro is
       dnl invoked! To detect that, one needs to look at every instance
       dnl of the macro expansion in the generated configure script.
       case $1 in
         @<:@\\/@:>@* | ?:@<:@\\/@:>@* @:}@ $2=$1 ;;
         *@:}@ $2="$PWD"/$1 ;;
       esac
    fi
    dnl Check for space in the directory
    if test `echo $1|cut -d' ' -f1` != $1 ; then
        AC_MSG_ERROR($1 directory shall not contain any space.)
    fi
 else
    AC_MSG_ERROR($1 shall be a valid directory)
 fi
])


dnl  MPFR_C_LONG_DOUBLE_FORMAT
dnl  -------------------------
dnl  Determine the format of a long double.
dnl
dnl  The object file is grepped, so as to work when cross compiling.  A
dnl  start and end sequence is included to avoid false matches, and
dnl  allowance is made for the desired data crossing an "od -b" line
dnl  boundary.  The test number is a small integer so it should appear
dnl  exactly, no rounding or truncation etc.
dnl
dnl  "od -b" is supported even by Unix V7, and the awk script used doesn't
dnl  have functions or anything, so even an "old" awk should suffice.
dnl
dnl  The 10-byte IEEE extended format is generally padded to either 12 or 16
dnl  bytes for alignment purposes.  The SVR4 i386 ABI is 12 bytes, or i386
dnl  gcc -m128bit-long-double selects 16 bytes.  IA-64 is 16 bytes in LP64
dnl  mode, or 12 bytes in ILP32 mode.  The first 10 bytes is the relevant
dnl  part in all cases (big and little endian).
dnl
dnl  Enhancements:
dnl
dnl  Could match more formats, but no need to worry until there's code
dnl  wanting to use them.
dnl
dnl  Don't want to duplicate the double matching from GMP_C_DOUBLE_FORMAT,
dnl  perhaps we should merge with that macro, to match data formats
dnl  irrespective of the C type in question.  Or perhaps just let the code
dnl  use DOUBLE macros when sizeof(double)==sizeof(long double).

AC_DEFUN([MPFR_C_LONG_DOUBLE_FORMAT],
[AC_REQUIRE([AC_PROG_CC])
AC_REQUIRE([AC_PROG_AWK])
AC_REQUIRE([AC_OBJEXT])
AC_CHECK_TYPES([long double])
AC_CACHE_CHECK([format of `long double' floating point],
                mpfr_cv_c_long_double_format,
[mpfr_cv_c_long_double_format=unknown
if test "$ac_cv_type_long_double" != yes; then
  mpfr_cv_c_long_double_format="not available"
else
  cat >conftest.c <<\EOF
[
/* "before" is 16 bytes to ensure there's no padding between it and "x".
   We're not expecting any "long double" bigger than 16 bytes or with
   alignment requirements stricter than 16 bytes.  */
struct {
  char         before[16];
  long double  x;
  char         after[8];
} foo = {
  { '\0', '\0', '\0', '\0', '\0', '\0', '\0', '\0',
    '\001', '\043', '\105', '\147', '\211', '\253', '\315', '\357' },
  -123456789.0,
  { '\376', '\334', '\272', '\230', '\166', '\124', '\062', '\020' }
};
]
EOF
  mpfr_compile="$CC $CFLAGS $CPPFLAGS -c conftest.c >&AS_MESSAGE_LOG_FD 2>&1"
  if AC_TRY_EVAL(mpfr_compile); then
    cat >conftest.awk <<\EOF
[
BEGIN {
  found = 0
}

# got[] holds a sliding window of bytes read the input.  got[0] is the most
# recent byte read, and got[31] the oldest byte read, so when looking to
# match some data the indices are "reversed".
#
{
  for (f = 2; f <= NF; f++)
    {
      # new byte, shift others up
      for (i = 31; i >= 0; i--)
        got[i+1] = got[i];
      got[0] = $f;

      # end sequence
      if (got[7] != "376") continue
      if (got[6] != "334") continue
      if (got[5] != "272") continue
      if (got[4] != "230") continue
      if (got[3] != "166") continue
      if (got[2] != "124") continue
      if (got[1] != "062") continue
      if (got[0] != "020") continue

      # start sequence, with 8-byte body
      if (got[23] == "001" && \
          got[22] == "043" && \
          got[21] == "105" && \
          got[20] == "147" && \
          got[19] == "211" && \
          got[18] == "253" && \
          got[17] == "315" && \
          got[16] == "357")
        {
          saw = " (" got[15] \
                 " " got[14] \
                 " " got[13] \
                 " " got[12] \
                 " " got[11] \
                 " " got[10] \
                 " " got[9]  \
                 " " got[8] ")"

          if (got[15] == "301" && \
              got[14] == "235" && \
              got[13] == "157" && \
              got[12] == "064" && \
              got[11] == "124" && \
              got[10] == "000" && \
              got[9] ==  "000" && \
              got[8] ==  "000")
            {
              print "IEEE double, big endian"
              found = 1
              exit
            }

          if (got[15] == "000" && \
              got[14] == "000" && \
              got[13] == "000" && \
              got[12] == "124" && \
              got[11] == "064" && \
              got[10] == "157" && \
              got[9] ==  "235" && \
              got[8] ==  "301")
            {
              print "IEEE double, little endian"
              found = 1
              exit
            }
        }

      # start sequence, with 12-byte body
      if (got[27] == "001" && \
          got[26] == "043" && \
          got[25] == "105" && \
          got[24] == "147" && \
          got[23] == "211" && \
          got[22] == "253" && \
          got[21] == "315" && \
          got[20] == "357")
        {
          saw = " (" got[19] \
                 " " got[18] \
                 " " got[17] \
                 " " got[16] \
                 " " got[15] \
                 " " got[14] \
                 " " got[13] \
                 " " got[12] \
                 " " got[11] \
                 " " got[10] \
                 " " got[9]  \
                 " " got[8] ")"

          if (got[19] == "000" && \
              got[18] == "000" && \
              got[17] == "000" && \
              got[16] == "000" && \
              got[15] == "240" && \
              got[14] == "242" && \
              got[13] == "171" && \
              got[12] == "353" && \
              got[11] == "031" && \
              got[10] == "300")
            {
              print "IEEE extended, little endian (12 bytes)"
              found = 1
              exit
            }

          if (got[19] == "300" && \
              got[18] == "031" && \
              got[17] == "000" && \
              got[16] == "000" && \
              got[15] == "353" && \
              got[14] == "171" && \
              got[13] == "242" && \
              got[12] == "240" && \
              got[11] == "000" && \
              got[10] == "000" && \
              got[09] == "000" && \
              got[08] == "000")
            {
              # format found on m68k
              print "IEEE extended, big endian (12 bytes)"
              found = 1
              exit
            }
        }

      # start sequence, with 16-byte body
      if (got[31] == "001" && \
          got[30] == "043" && \
          got[29] == "105" && \
          got[28] == "147" && \
          got[27] == "211" && \
          got[26] == "253" && \
          got[25] == "315" && \
          got[24] == "357")
        {
          saw = " (" got[23] \
                 " " got[22] \
                 " " got[21] \
                 " " got[20] \
                 " " got[19] \
                 " " got[18] \
                 " " got[17] \
                 " " got[16] \
                 " " got[15] \
                 " " got[14] \
                 " " got[13] \
                 " " got[12] \
                 " " got[11] \
                 " " got[10] \
                 " " got[9]  \
                 " " got[8] ")"

          if (got[23] == "000" && \
              got[22] == "000" && \
              got[21] == "000" && \
              got[20] == "000" && \
              got[19] == "240" && \
              got[18] == "242" && \
              got[17] == "171" && \
              got[16] == "353" && \
              got[15] == "031" && \
              got[14] == "300")
            {
              print "IEEE extended, little endian (16 bytes)"
              found = 1
              exit
            }

          if (got[23] == "300" && \
              got[22] == "031" && \
              got[21] == "326" && \
              got[20] == "363" && \
              got[19] == "105" && \
              got[18] == "100" && \
              got[17] == "000" && \
              got[16] == "000" && \
              got[15] == "000" && \
              got[14] == "000" && \
              got[13] == "000" && \
              got[12] == "000" && \
              got[11] == "000" && \
              got[10] == "000" && \
              got[9]  == "000" && \
              got[8]  == "000")
            {
              # format used on HP 9000/785 under HP-UX
              print "IEEE quad, big endian"
              found = 1
              exit
            }

          if (got[23] == "000" && \
              got[22] == "000" && \
              got[21] == "000" && \
              got[20] == "000" && \
              got[19] == "000" && \
              got[18] == "000" && \
              got[17] == "000" && \
              got[16] == "000" && \
              got[15] == "000" && \
              got[14] == "000" && \
              got[13] == "100" && \
              got[12] == "105" && \
              got[11] == "363" && \
              got[10] == "326" && \
              got[9]  == "031" && \
	      got[8]  == "300")
            {
              print "IEEE quad, little endian"
              found = 1
              exit
            }

          if (got[23] == "301" && \
              got[22] == "235" && \
              got[21] == "157" && \
              got[20] == "064" && \
              got[19] == "124" && \
              got[18] == "000" && \
              got[17] == "000" && \
              got[16] == "000" && \
              got[15] == "000" && \
              got[14] == "000" && \
              got[13] == "000" && \
              got[12] == "000" && \
              got[11] == "000" && \
              got[10] == "000" && \
              got[9]  == "000" && \
              got[8]  == "000")
            {
              # format used on 32-bit PowerPC (Mac OS X and Debian GNU/Linux)
              print "possibly double-double, big endian"
              found = 1
              exit
            }

          if (got[23] == "000" && \
              got[22] == "000" && \
              got[21] == "000" && \
              got[20] == "124" && \
              got[19] == "064" && \
              got[18] == "157" && \
              got[17] == "235" && \
              got[16] == "301" && \
              got[15] == "000" && \
              got[14] == "000" && \
              got[13] == "000" && \
              got[12] == "000" && \
              got[11] == "000" && \
              got[10] == "000" && \
              got[9]  == "000" && \
              got[8]  == "000")
            {
              # format used on ppc64le
              print "possibly double-double, little endian"
              found = 1
              exit
            }
        }
    }
}

END {
  if (! found)
    print "unknown", saw
}
]
EOF
    mpfr_cv_c_long_double_format=`od -b conftest.$OBJEXT | $AWK -f conftest.awk`
    case $mpfr_cv_c_long_double_format in
    unknown*)
      echo "cannot match anything, conftest.$OBJEXT contains" >&AS_MESSAGE_LOG_FD
      od -b conftest.$OBJEXT >&AS_MESSAGE_LOG_FD
      ;;
    esac
  else
    AC_MSG_WARN([oops, cannot compile test program])
  fi
fi
rm -f conftest*
])

AH_VERBATIM([HAVE_LDOUBLE],
[/* Define one of the following to 1 for the format of a `long double'.
   If your format is not among these choices, or you don't know what it is,
   then leave all undefined.
   IEEE_EXT is the 10-byte IEEE extended precision format.
   IEEE_QUAD is the 16-byte IEEE quadruple precision format.
   LITTLE or BIG is the endianness.  */
#undef HAVE_LDOUBLE_IEEE_EXT_LITTLE
#undef HAVE_LDOUBLE_IEEE_QUAD_BIG])

case $mpfr_cv_c_long_double_format in
  "IEEE double, big endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IS_DOUBLE, 1)
    ;;
  "IEEE double, little endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IS_DOUBLE, 1)
    ;;
  "IEEE extended, little endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IEEE_EXT_LITTLE, 1)
    ;;
  "IEEE extended, big endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IEEE_EXT_BIG, 1)
    ;;
  "IEEE quad, big endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IEEE_QUAD_BIG, 1)
    ;;
  "IEEE quad, little endian"*)
    AC_DEFINE(HAVE_LDOUBLE_IEEE_QUAD_LITTLE, 1)
    ;;
  "possibly double-double, big endian"*)
    AC_MSG_WARN([This format is known on GCC/PowerPC platforms,])
    AC_MSG_WARN([but due to GCC PR26374, we can't test further.])
    AC_MSG_WARN([You can safely ignore this warning, though.])
    AC_DEFINE(HAVE_LDOUBLE_MAYBE_DOUBLE_DOUBLE, 1)
    ;;
  "possibly double-double, little endian"*)
    AC_MSG_WARN([This format is known on GCC/PowerPC platforms,])
    AC_MSG_WARN([but due to GCC PR26374, we can't test further.])
    AC_MSG_WARN([You can safely ignore this warning, though.])
    AC_DEFINE(HAVE_LDOUBLE_MAYBE_DOUBLE_DOUBLE, 1)
    ;;
  unknown* | "not available"*)
    ;;
  *)
    AC_MSG_WARN([unrecognized long double FP format: $mpfr_cv_c_long_double_format])
    ;;
esac
])

dnl  MPFR_CHECK_LIBM
dnl  ---------------
dnl  Determine a math library -lm to use.

AC_DEFUN([MPFR_CHECK_LIBM],
[AC_REQUIRE([AC_CANONICAL_HOST])
AC_SUBST(MPFR_LIBM,'')
case $host in
  *-*-beos* | *-*-cygwin* | *-*-pw32*)
    # According to libtool AC CHECK LIBM, these systems don't have libm
    ;;
  *-*-solaris*)
    # On Solaris the math functions new in C99 are in -lm9x.
    # FIXME: Do we need -lm9x as well as -lm, or just instead of?
    AC_CHECK_LIB(m9x, main, MPFR_LIBM="-lm9x")
    AC_CHECK_LIB(m,   main, MPFR_LIBM="$MPFR_LIBM -lm")
    ;;
  *-ncr-sysv4.3*)
    # FIXME: What does -lmw mean?  Libtool AC CHECK LIBM does it this way.
    AC_CHECK_LIB(mw, _mwvalidcheckl, MPFR_LIBM="-lmw")
    AC_CHECK_LIB(m, main, MPFR_LIBM="$MPFR_LIBM -lm")
    ;;
  *)
    AC_CHECK_LIB(m, main, MPFR_LIBM="-lm")
    ;;
esac
])

dnl  MPFR_CHECK_LIBQUADMATH
dnl  ---------------
dnl  Determine a math library -lquadmath to use.
AC_DEFUN([MPFR_CHECK_LIBQUADMATH],
[AC_REQUIRE([AC_CANONICAL_HOST])
AC_SUBST(MPFR_LIBQUADMATH,'')
case $host in
  *)
    AC_CHECK_LIB(quadmath, main, MPFR_LIBQUADMATH="-lquadmath")
    ;;
esac
])

dnl  MPFR_LD_SEARCH_PATHS_FIRST
dnl  --------------------------

AC_DEFUN([MPFR_LD_SEARCH_PATHS_FIRST],
[case "$LD $LDFLAGS" in
  *-Wl,-search_paths_first*) ;;
  *) AC_MSG_CHECKING([if the compiler understands -Wl,-search_paths_first])
     saved_LDFLAGS="$LDFLAGS"
     LDFLAGS="-Wl,-search_paths_first $LDFLAGS"
     AC_LINK_IFELSE([AC_LANG_PROGRAM([[]], [[]])],
       [AC_MSG_RESULT(yes)],
       [AC_MSG_RESULT(no)]
        LDFLAGS="$saved_LDFLAGS")
     ;;
 esac
])


dnl  GMP_C_ATTRIBUTE_MODE
dnl  --------------------
dnl  Introduced in gcc 2.2, but perhaps not in all Apple derived versions.
dnl  Needed for mpfr-longlong.h; this is currently necessary for s390.
dnl
dnl  TODO: Replace this with a cleaner type size detection, as this
dnl  solution only works with gcc and assumes CHAR_BIT == 8. Probably use
dnl  <stdint.h>, and <https://gcc.gnu.org/viewcvs/gcc/trunk/config/stdint.m4>
dnl  as a fallback.

AC_DEFUN([GMP_C_ATTRIBUTE_MODE],
[AC_CACHE_CHECK([whether gcc __attribute__ ((mode (XX))) works],
               gmp_cv_c_attribute_mode,
[AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[typedef int SItype __attribute__ ((mode (SI)));]], [[]])],
 gmp_cv_c_attribute_mode=yes, gmp_cv_c_attribute_mode=no)
])
if test $gmp_cv_c_attribute_mode = yes; then
 AC_DEFINE(HAVE_ATTRIBUTE_MODE, 1,
 [Define to 1 if the compiler accepts gcc style __attribute__ ((mode (XX)))])
fi
])


dnl  MPFR_FUNC_GMP_PRINTF_SPEC
dnl  ------------------------------------
dnl  MPFR_FUNC_GMP_PRINTF_SPEC(spec, type, [includes], [if-true], [if-false])
dnl  Check if gmp_sprintf supports the conversion specification 'spec'
dnl  with type 'type'.
dnl  Expand 'if-true' if printf supports 'spec', 'if-false' otherwise.

AC_DEFUN([MPFR_FUNC_GMP_PRINTF_SPEC],[
AC_MSG_CHECKING(if gmp_printf supports "%$1")
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <stdio.h>
#include <string.h>
$3
#include <gmp.h>
]], [[
  char s[256];
  $2 a = 17;

  if (gmp_sprintf (s, "(%0.0$1)(%d)", a, 42) == -1) return 1;
  return (strcmp (s, "(17)(42)") != 0);
]])],
  [AC_MSG_RESULT(yes)
  $4],
  [AC_MSG_RESULT(no)
  $5],
  [AC_MSG_RESULT(cross-compiling, assuming yes)
  $4])
])


dnl MPFR_CHECK_PRINTF_SPEC
dnl ----------------------
dnl Check if gmp_printf supports some optional length modifiers.
dnl Defined symbols are negative to shorten the gcc command line.

AC_DEFUN([MPFR_CHECK_PRINTF_SPEC], [
AC_REQUIRE([MPFR_CONFIGS])dnl
if test "$ac_cv_type_intmax_t" = yes; then
 MPFR_FUNC_GMP_PRINTF_SPEC([jd], [intmax_t], [
#ifdef HAVE_INTTYPES_H
# include <inttypes.h>
#endif
#ifdef HAVE_STDINT_H
# include <stdint.h>
#endif
         ],,
         [AC_DEFINE([NPRINTF_J], 1, [gmp_printf cannot read intmax_t])])
fi

MPFR_FUNC_GMP_PRINTF_SPEC([hhd], [char], [
#include <gmp.h>
         ],,
         [AC_DEFINE([NPRINTF_HH], 1, [gmp_printf cannot use `hh' length modifier])])

MPFR_FUNC_GMP_PRINTF_SPEC([lld], [long long int], [
#include <gmp.h>
         ],,
         [AC_DEFINE([NPRINTF_LL], 1, [gmp_printf cannot read long long int])])

MPFR_FUNC_GMP_PRINTF_SPEC([Lf], [long double], [
#include <gmp.h>
         ],
         [AC_DEFINE([PRINTF_L], 1, [gmp_printf can read long double])],
         [AC_DEFINE([NPRINTF_L], 1, [gmp_printf cannot read long double])])

MPFR_FUNC_GMP_PRINTF_SPEC([td], [ptrdiff_t], [
#if defined (__cplusplus)
#include <cstddef>
#else
#include <stddef.h>
#endif
#include <gmp.h>
    ],
    [AC_DEFINE([PRINTF_T], 1, [gmp_printf can read ptrdiff_t])],
    [AC_DEFINE([NPRINTF_T], 1, [gmp_printf cannot read ptrdiff_t])])
])

dnl MPFR_CHECK_PRINTF_GROUPFLAG
dnl ---------------------------
dnl Check the support of the group flag for native integers, which is
dnl a Single UNIX Specification extension.
dnl This will be used to enable some tests, as the implementation of
dnl the P length modifier for mpfr_*printf relies on this support.

AC_DEFUN([MPFR_CHECK_PRINTF_GROUPFLAG], [
AC_MSG_CHECKING(if gmp_printf supports the ' group flag)
AC_RUN_IFELSE([AC_LANG_PROGRAM([[
#include <string.h>
#include <gmp.h>
]], [[
  char s[256];

  if (gmp_sprintf (s, "%'d", 17) == -1) return 1;
  return (strcmp (s, "17") != 0);
]])],
  [AC_MSG_RESULT(yes)
   AC_DEFINE([PRINTF_GROUPFLAG], 1, [Define if gmp_printf supports the ' group flag])],
  [AC_MSG_RESULT(no)],
  [AC_MSG_RESULT(cannot test, assume no)])
])
])

dnl MPFR_LTO
dnl --------
dnl To be representative, we need:
dnl * to compile a source,
dnl * to generate a library archive,
dnl * to generate a program with this archive.
AC_DEFUN([MPFR_LTO],[
dnl Check for -flto support
CFLAGS="$CFLAGS -flto"

AC_MSG_CHECKING([if Link Time Optimisation flag '-flto' is supported...])
AC_COMPILE_IFELSE([AC_LANG_SOURCE([[
int main(void) { return 0; }
  ]])],
     [AC_MSG_RESULT(yes)
     ],
     [AC_MSG_RESULT(no)
      AC_MSG_ERROR([Link Time Optimisation flag '-flto' is not supported.])
     ])

dnl Check if it works...
mpfr_compile_and_link()
{
   echo "int f(int); int f(int n) { return n; }" > conftest-f.c
   echo "int f(int); int main() { return f(0); }" > conftest-m.c
   echo "$CC $CFLAGS -c -o conftest-f.o conftest-f.c" >&2
   $CC $CFLAGS -c -o conftest-f.o conftest-f.c || return 1
   echo "$AR cru conftest-lib.a conftest-f.o" >&2
   $AR cru conftest-lib.a conftest-f.o || return 1
   echo "$RANLIB conftest-lib.a" >&2
   $RANLIB conftest-lib.a || return 1
   echo "$CC $CFLAGS conftest-m.c conftest-lib.a" >&2
   $CC $CFLAGS conftest-m.c conftest-lib.a || return 1
   return 0
}
   AC_MSG_CHECKING([if Link Time Optimisation works with AR=$AR])
   if mpfr_compile_and_link 2> conftest-log1.txt ; then
      cat conftest-log1.txt >&AS_MESSAGE_LOG_FD
      AC_MSG_RESULT(yes)
   else
      cat conftest-log1.txt >&AS_MESSAGE_LOG_FD
      AC_MSG_RESULT(no)
      AR=gcc-ar
      RANLIB=gcc-ranlib
      AC_MSG_CHECKING([if Link Time Optimisation works with AR=$AR])
      if mpfr_compile_and_link 2> conftest-log2.txt; then
         cat conftest-log2.txt >&AS_MESSAGE_LOG_FD
         AC_MSG_RESULT(yes)
      else
        cat conftest-log2.txt >&AS_MESSAGE_LOG_FD
        AC_MSG_RESULT(no)
        AC_MSG_ERROR([Link Time Optimisation is not supported (see config.log for details).])
      fi
   fi
rm -f conftest*
])

dnl MPFR_CHECK_SHARED_CACHE
dnl ----------------------
dnl Check if the conditions for the shared cache are met:
dnl  * either pthread / C11 are available.
dnl  * either constructor or once.
AC_DEFUN([MPFR_CHECK_SHARED_CACHE], [
  AC_MSG_CHECKING(if shared cache is supported)
  if test "$enable_logging" = yes ; then
    AC_MSG_RESULT(no)
    AC_MSG_ERROR([shared cache does not work with logging support.])
dnl because logging support disables threading support
  elif test "$enable_thread_safe" != yes ; then
    AC_MSG_RESULT(no)
    AC_MSG_ERROR([shared cache needs thread attribute.])
  elif test "$ax_pthread_ok" != yes && "$mpfr_c11_thread_ok" != yes ; then
    AC_MSG_RESULT(no)
    AC_MSG_ERROR([shared cache needs pthread/C11 library.])
  else
    AC_MSG_RESULT(yes)
    if test "$ax_pthread_ok" = yes ; then
        CFLAGS="$CFLAGS $PTHREAD_CFLAGS"
        LIBS="$LIBS $PTHREAD_LIBS"
    fi
  fi
])

dnl MPFR_CHECK_CONSTRUCTOR_ATTR
dnl ---------------------------
dnl Check for constructor/destructor attributes to function.
dnl Output: Define
dnl  * MPFR_HAVE_CONSTRUCTOR_ATTR C define
dnl  * mpfr_have_constructor_destructor_attributes shell variable to yes
dnl if supported.
AC_DEFUN([MPFR_CHECK_CONSTRUCTOR_ATTR], [
AC_MSG_CHECKING([for constructor and destructor attributes])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <stdlib.h>
int x = 0;
__attribute__((constructor)) static void
call_f(void) { x = 1742; }
__attribute__((destructor)) static void
call_g(void) { x = 1448; }
]], [[
    return (x == 1742) ? 0 : 1;
]])], [AC_MSG_RESULT(yes)
       mpfr_have_constructor_destructor_attributes=yes ],
      [AC_MSG_RESULT(no)]
)

if test "$mpfr_have_constructor_destructor_attributes" = "yes"; then
  AC_DEFINE(MPFR_HAVE_CONSTRUCTOR_ATTR, 1,
   [Define if the constructor/destructor GCC attributes are supported.])
fi
])

dnl MPFR_CHECK_C11_THREAD
dnl ---------------------
dnl Check for C11 thread support
dnl Output: Define
dnl  * MPFR_HAVE_C11_LOCK C define
dnl  * mpfr_c11_thread_ok shell variable to yes
dnl if supported.
dnl We don't check for __STDC_NO_THREADS__ define variable but rather try to mimic our usage.
AC_DEFUN([MPFR_CHECK_C11_THREAD], [
AC_MSG_CHECKING([for ISO C11 thread support])
AC_LINK_IFELSE([AC_LANG_PROGRAM([[
#include <assert.h>
#include <threads.h>
 mtx_t lock;
 once_flag once = ONCE_FLAG_INIT;
 thrd_t thd_idx;
 int x = 0;
 void once_call (void) { x = 1; }
]], [[
    int err;
    err = mtx_init(&lock, mtx_plain);
    assert(err == thrd_success);
    err = mtx_lock(&lock);
    assert(err == thrd_success);
    err = mtx_unlock(&lock);
    assert(err == thrd_success);
    mtx_destroy(&lock);
    once_call(&once, once_call);
    return x == 1 ? 0 : -1;
]])], [AC_MSG_RESULT(yes)
       mpfr_c11_thread_ok=yes ],
      [AC_MSG_RESULT(no)]
)

if test "$mpfr_c11_thread_ok" = "yes"; then
  AC_DEFINE(MPFR_HAVE_C11_LOCK, 1,
   [Define if the ISO C11 Thread is supported.])
fi
])