It seems that we need prototypes for long double math functions, even
when !__HAVE_LONG_DOUBLE -- they still exist, but just aren't...

This might unbreak several more builds.

This set of changes seem to be sufficient to make at least the
amd64 port build again after the long double math function infusion
from yesterday.

Feel free (without consulting me, I'm just hacking about) to revert
any of this and do it a better way.

Note: this is not intended to be complete, in particular, in <math.h>
I added prototypes only for the functions that are needed for a
successful build, not all of the new ones...

Add sincos{,f,l} from FreeBSD

Correct C99 / C++11 feature testing. Document remaining missing symbols.

branches: 1.65.2; 1.65.4;
Fix the rest of the functions that have been made visible by xopen5,
posix 2017, and c99.

PR/53393: coypu: move rint to the right section.

branches: 1.63.24;
Allow archs to provide __isinf() and __isnan() as inline functions instead
of macros.

If __FLT_EVAL_METHOD__ exists and is in {0,1,2}, provide float_t and
double_t.

FP_ILOGBNAN should be INT_MAX

Define FP_ILOGB0 and FP_ILOGBNAN

Add missing long double functions

branches: 1.58.2;
oops, forgot to commit: add tgamma and tgammaf

branches: 1.57.2;
Use __INFINITY in preference to __builtin_inff(). Currently only VAX defines
__INFINITY and allows to use __builtin_huge_valf() instead since gcc4.5 will
error out unconditionally if __builtin_inf() is used.

branches: 1.56.2;
Add ilogbl(3).

Add scalbnl to libm.

remquo{,f} from FreeBSD via Stathis Kamperis

branches: 1.53.2;
Commit SoC long double support from Stathis Kamperis

typo in comment

PR 36988: ISO C99 mandates that HUGE_VAL, HUGE_VALF, HUGE_VALL and
INFINITY are constant expressions.

Use the GCC builtins if possible to implement them.

branches: 1.50.2;
Add exp2 and exp2m

oops forgot to commit that one--- Add f{dim,min,max}{,l,f}

Make NAN a compile time constant (with help from the compiler). Suggested
by krister. Fixes PR 40695.
Make references to the old construct (which we can't remove for binary
compatibility reasons) emit a linker warning.

branches: 1.47.8; 1.47.10; 1.47.12; 1.47.16;
make name const.

branches: 1.46.10;
-get cabs() and cabsf() out of public view - these are historical and
conflict with C99 functions which are builtins in newer gcc
(actually, the old cabs() is ABI compatible with the new _complex one
on i386, but this is purely accidental)
remove public prototypes and manpages, move the code into a compat
subdirectory as libc does so that binary compatibility is kept
-add a manpage for the isgreater() etc macros, borrowed from FreeBSD

add the isgreater() at al. comparision macros defined in C99 and SUSv3;
use just the primitive macros for now (identical to FreeBSD/DragonFly)
which don't use gcc internals, the rest can go in after some testing;
addresses PR standards/25520

Add trunc() and truncf() C99 functions, from FreeBSD.

Add nan(3), nanf(3), and nanl(3); welcome libm.so.0.4.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Add log2f

de-__P -- the hack is long since useless. Discussed with christos,
matt, kleink, others. Approved by christos.

Move round() to the C99 section.

lib/24252: Add C99 functions round(3) and roundf(3).
libm minor unchanged; ride a recent bump.

From FreeBSD.

being here, add the (l)lround(f) functions to libm
(almost the same like lrint)

shut up lint's "long long" warnings

Add (l)lrint(f).
Being here, move some C99 function declarations into the appropriate
section (as discussed with kleink).

branches: 1.34.2;
* Turn isinf(3) and isnan(3) into C99-style macros.
* Make it possible for ports to override these (i.e., VAX).
* Remove isnanl(), which was internal to libc only.

Make lint happy for __fpmacro_unary_floating() users.

Move FP_ROP and FP_DIRTYZERO from <math.h> to <vax/math.h> (where they
would have ended up if I hadn't missed the latter file in the commit).

Add FP_ROP and FP_DIRTYZERO to MD range for VAX.

Close the comment after #endif.

Add C99 fpclassify(), isfinite(), isnormal(), and signbit() macros.

C99 7.12#4: provide INFINITY, which defaults to HUGE_VALF.
If infinities are not available, the machine-dependent header must define
__INFINITY as a positive constant of type float that overflows.

C99: provide HUGE_VALF and HUGE_VALL.

Add long double support.

Add a new feature-test macro, _NETBSD_SOURCE. If this is defined
by the application, all NetBSD interfaces are made visible, even
if some other feature-test macro (like _POSIX_C_SOURCE) is defined.
<sys/featuretest.h> defined _NETBSD_SOURCE if none of _ANSI_SOURCE,
_POSIX_C_SOURCE and _XOPEN_SOURCE is defined, so as to preserve
existing behaviour.

This has two major advantages:
+ Programs that require non-POSIX facilities but define _POSIX_C_SOURCE
can trivially be overruled by putting -D_NETBSD_SOURCE in their CFLAGS.
+ It makes most of the #ifs simpler, in that they're all now ORs of the
various macros, rather than having checks for (!defined(_ANSI_SOURCE) ||
!defined(_POSIX_C_SOURCE) || !defined(_XOPEN_SOURCE)) all over the place.

I've tried not to change the semantics of the headers in any case where
_NETBSD_SOURCE wasn't defined, but there were some places where the
current semantics were clearly mad, and retaining them was harder than
correcting them. In particular, I've mostly normalised things so that
_ANSI_SOURCE gets you the smallest set of stuff, then _POSIX_C_SOURCE,
_XOPEN_SOURCE and _NETBSD_SOURCE in that order.

Tested by building for vax, encouraged by thorpej, and uncontested in
tech-userlevel for a week.

Clean up some rampant code duplication wrt ieee number handling:
- Add alignment-safe double and float unions.
- Use the above for the __infinity and __nan constants on all
architectures that use the standard ieee754 representation of
those constants.
- Add a single copy of various ieee754 math functions (frexp, isinf,
isnan, ldexp and modf) that had numerous duplicates among the
arch-specific directories.
- Use the above functions on all architectures where the generic C
versions where used. Architectures that had local assembly
routines are untouched (for those functions only).

branches: 1.23.2;
protect the incomplete cabs declarations with #ifndef __MATH_PRIVATE__ so that
the code that defines the functions can declare them properly.

Remove 'extern' from function declarations.

branches: 1.21.2;
const -> __const and include <sys/cdefs.h> earlier; fixes PR lib/9052
by Takahiro Kambe.

C99: Define a NAN macro in <math.h> which evaulates to a constant expression of
a single-precision quiet NaN; only to be defined on platforms that do support
this value.

Add missing isinff() declaration.

This file has still missing prototypes: cabs and cabsf... At least comment
what the arguments should be.

s/void/const void/

cast to void * first to avoid lint warning.

Oops, isinf() was never included in any X/Open interface specification.

const poisoning.

Fix comment typo in previous.

Reorganize name space protection.

C++ has a very different concept of "exception." Delimit struct exception
and matherr() with #ifndef __cplusplus.

new RCS ID format.

Added prototypes for float versions of math functions.
Removed prototypes for internal functions.

Add cabs() and drem(), for BSD libm compatibility.

Add missing _. reported by several folks.

New math.h, based on fdlibm's fdlibm.h.

Moved definition of M_FOO constants into !_ANSI_SOURCE && !_POSIX_SOURCE
conditional.

Use __pure qualifier.

update for better FP routines, from AT&T & elsewhere

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision

Add sincos{,f,l} from FreeBSD

Correct C99 / C++11 feature testing. Document remaining missing symbols.

branches: 1.65.2; 1.65.4;
Fix the rest of the functions that have been made visible by xopen5,
posix 2017, and c99.

PR/53393: coypu: move rint to the right section.

branches: 1.63.24;
Allow archs to provide __isinf() and __isnan() as inline functions instead
of macros.

If __FLT_EVAL_METHOD__ exists and is in {0,1,2}, provide float_t and
double_t.

FP_ILOGBNAN should be INT_MAX

Define FP_ILOGB0 and FP_ILOGBNAN

Add missing long double functions

branches: 1.58.2;
oops, forgot to commit: add tgamma and tgammaf

branches: 1.57.2;
Use __INFINITY in preference to __builtin_inff(). Currently only VAX defines
__INFINITY and allows to use __builtin_huge_valf() instead since gcc4.5 will
error out unconditionally if __builtin_inf() is used.

branches: 1.56.2;
Add ilogbl(3).

Add scalbnl to libm.

remquo{,f} from FreeBSD via Stathis Kamperis

branches: 1.53.2;
Commit SoC long double support from Stathis Kamperis

typo in comment

PR 36988: ISO C99 mandates that HUGE_VAL, HUGE_VALF, HUGE_VALL and
INFINITY are constant expressions.

Use the GCC builtins if possible to implement them.

branches: 1.50.2;
Add exp2 and exp2m

oops forgot to commit that one--- Add f{dim,min,max}{,l,f}

Make NAN a compile time constant (with help from the compiler). Suggested
by krister. Fixes PR 40695.
Make references to the old construct (which we can't remove for binary
compatibility reasons) emit a linker warning.

branches: 1.47.8; 1.47.10; 1.47.12; 1.47.16;
make name const.

branches: 1.46.10;
-get cabs() and cabsf() out of public view - these are historical and
conflict with C99 functions which are builtins in newer gcc
(actually, the old cabs() is ABI compatible with the new _complex one
on i386, but this is purely accidental)
remove public prototypes and manpages, move the code into a compat
subdirectory as libc does so that binary compatibility is kept
-add a manpage for the isgreater() etc macros, borrowed from FreeBSD

add the isgreater() at al. comparision macros defined in C99 and SUSv3;
use just the primitive macros for now (identical to FreeBSD/DragonFly)
which don't use gcc internals, the rest can go in after some testing;
addresses PR standards/25520

Add trunc() and truncf() C99 functions, from FreeBSD.

Add nan(3), nanf(3), and nanl(3); welcome libm.so.0.4.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Add log2f

de-__P -- the hack is long since useless. Discussed with christos,
matt, kleink, others. Approved by christos.

Move round() to the C99 section.

lib/24252: Add C99 functions round(3) and roundf(3).
libm minor unchanged; ride a recent bump.

From FreeBSD.

being here, add the (l)lround(f) functions to libm
(almost the same like lrint)

shut up lint's "long long" warnings

Add (l)lrint(f).
Being here, move some C99 function declarations into the appropriate
section (as discussed with kleink).

branches: 1.34.2;
* Turn isinf(3) and isnan(3) into C99-style macros.
* Make it possible for ports to override these (i.e., VAX).
* Remove isnanl(), which was internal to libc only.

Make lint happy for __fpmacro_unary_floating() users.

Move FP_ROP and FP_DIRTYZERO from <math.h> to <vax/math.h> (where they
would have ended up if I hadn't missed the latter file in the commit).

Add FP_ROP and FP_DIRTYZERO to MD range for VAX.

Close the comment after #endif.

Add C99 fpclassify(), isfinite(), isnormal(), and signbit() macros.

C99 7.12#4: provide INFINITY, which defaults to HUGE_VALF.
If infinities are not available, the machine-dependent header must define
__INFINITY as a positive constant of type float that overflows.

C99: provide HUGE_VALF and HUGE_VALL.

Add long double support.

Add a new feature-test macro, _NETBSD_SOURCE. If this is defined
by the application, all NetBSD interfaces are made visible, even
if some other feature-test macro (like _POSIX_C_SOURCE) is defined.
<sys/featuretest.h> defined _NETBSD_SOURCE if none of _ANSI_SOURCE,
_POSIX_C_SOURCE and _XOPEN_SOURCE is defined, so as to preserve
existing behaviour.

This has two major advantages:
+ Programs that require non-POSIX facilities but define _POSIX_C_SOURCE
can trivially be overruled by putting -D_NETBSD_SOURCE in their CFLAGS.
+ It makes most of the #ifs simpler, in that they're all now ORs of the
various macros, rather than having checks for (!defined(_ANSI_SOURCE) ||
!defined(_POSIX_C_SOURCE) || !defined(_XOPEN_SOURCE)) all over the place.

I've tried not to change the semantics of the headers in any case where
_NETBSD_SOURCE wasn't defined, but there were some places where the
current semantics were clearly mad, and retaining them was harder than
correcting them. In particular, I've mostly normalised things so that
_ANSI_SOURCE gets you the smallest set of stuff, then _POSIX_C_SOURCE,
_XOPEN_SOURCE and _NETBSD_SOURCE in that order.

Tested by building for vax, encouraged by thorpej, and uncontested in
tech-userlevel for a week.

Clean up some rampant code duplication wrt ieee number handling:
- Add alignment-safe double and float unions.
- Use the above for the __infinity and __nan constants on all
architectures that use the standard ieee754 representation of
those constants.
- Add a single copy of various ieee754 math functions (frexp, isinf,
isnan, ldexp and modf) that had numerous duplicates among the
arch-specific directories.
- Use the above functions on all architectures where the generic C
versions where used. Architectures that had local assembly
routines are untouched (for those functions only).

branches: 1.23.2;
protect the incomplete cabs declarations with #ifndef __MATH_PRIVATE__ so that
the code that defines the functions can declare them properly.

Remove 'extern' from function declarations.

branches: 1.21.2;
const -> __const and include <sys/cdefs.h> earlier; fixes PR lib/9052
by Takahiro Kambe.

C99: Define a NAN macro in <math.h> which evaulates to a constant expression of
a single-precision quiet NaN; only to be defined on platforms that do support
this value.

Add missing isinff() declaration.

This file has still missing prototypes: cabs and cabsf... At least comment
what the arguments should be.

s/void/const void/

cast to void * first to avoid lint warning.

Oops, isinf() was never included in any X/Open interface specification.

const poisoning.

Fix comment typo in previous.

Reorganize name space protection.

C++ has a very different concept of "exception." Delimit struct exception
and matherr() with #ifndef __cplusplus.

new RCS ID format.

Added prototypes for float versions of math functions.
Removed prototypes for internal functions.

Add cabs() and drem(), for BSD libm compatibility.

Add missing _. reported by several folks.

New math.h, based on fdlibm's fdlibm.h.

Moved definition of M_FOO constants into !_ANSI_SOURCE && !_POSIX_SOURCE
conditional.

Use __pure qualifier.

update for better FP routines, from AT&T & elsewhere

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision

Correct C99 / C++11 feature testing. Document remaining missing symbols.

Fix the rest of the functions that have been made visible by xopen5,
posix 2017, and c99.

PR/53393: coypu: move rint to the right section.

branches: 1.63.24;
Allow archs to provide __isinf() and __isnan() as inline functions instead
of macros.

If __FLT_EVAL_METHOD__ exists and is in {0,1,2}, provide float_t and
double_t.

FP_ILOGBNAN should be INT_MAX

Define FP_ILOGB0 and FP_ILOGBNAN

Add missing long double functions

branches: 1.58.2;
oops, forgot to commit: add tgamma and tgammaf

branches: 1.57.2;
Use __INFINITY in preference to __builtin_inff(). Currently only VAX defines
__INFINITY and allows to use __builtin_huge_valf() instead since gcc4.5 will
error out unconditionally if __builtin_inf() is used.

branches: 1.56.2;
Add ilogbl(3).

Add scalbnl to libm.

remquo{,f} from FreeBSD via Stathis Kamperis

branches: 1.53.2;
Commit SoC long double support from Stathis Kamperis

typo in comment

PR 36988: ISO C99 mandates that HUGE_VAL, HUGE_VALF, HUGE_VALL and
INFINITY are constant expressions.

Use the GCC builtins if possible to implement them.

branches: 1.50.2;
Add exp2 and exp2m

oops forgot to commit that one--- Add f{dim,min,max}{,l,f}

Make NAN a compile time constant (with help from the compiler). Suggested
by krister. Fixes PR 40695.
Make references to the old construct (which we can't remove for binary
compatibility reasons) emit a linker warning.

branches: 1.47.8; 1.47.10; 1.47.12; 1.47.16;
make name const.

branches: 1.46.10;
-get cabs() and cabsf() out of public view - these are historical and
conflict with C99 functions which are builtins in newer gcc
(actually, the old cabs() is ABI compatible with the new _complex one
on i386, but this is purely accidental)
remove public prototypes and manpages, move the code into a compat
subdirectory as libc does so that binary compatibility is kept
-add a manpage for the isgreater() etc macros, borrowed from FreeBSD

add the isgreater() at al. comparision macros defined in C99 and SUSv3;
use just the primitive macros for now (identical to FreeBSD/DragonFly)
which don't use gcc internals, the rest can go in after some testing;
addresses PR standards/25520

Add trunc() and truncf() C99 functions, from FreeBSD.

Add nan(3), nanf(3), and nanl(3); welcome libm.so.0.4.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Add log2f

de-__P -- the hack is long since useless. Discussed with christos,
matt, kleink, others. Approved by christos.

Move round() to the C99 section.

lib/24252: Add C99 functions round(3) and roundf(3).
libm minor unchanged; ride a recent bump.

From FreeBSD.

being here, add the (l)lround(f) functions to libm
(almost the same like lrint)

shut up lint's "long long" warnings

Add (l)lrint(f).
Being here, move some C99 function declarations into the appropriate
section (as discussed with kleink).

branches: 1.34.2;
* Turn isinf(3) and isnan(3) into C99-style macros.
* Make it possible for ports to override these (i.e., VAX).
* Remove isnanl(), which was internal to libc only.

Make lint happy for __fpmacro_unary_floating() users.

Move FP_ROP and FP_DIRTYZERO from <math.h> to <vax/math.h> (where they
would have ended up if I hadn't missed the latter file in the commit).

Add FP_ROP and FP_DIRTYZERO to MD range for VAX.

Close the comment after #endif.

Add C99 fpclassify(), isfinite(), isnormal(), and signbit() macros.

C99 7.12#4: provide INFINITY, which defaults to HUGE_VALF.
If infinities are not available, the machine-dependent header must define
__INFINITY as a positive constant of type float that overflows.

C99: provide HUGE_VALF and HUGE_VALL.

Add long double support.

Add a new feature-test macro, _NETBSD_SOURCE. If this is defined
by the application, all NetBSD interfaces are made visible, even
if some other feature-test macro (like _POSIX_C_SOURCE) is defined.
<sys/featuretest.h> defined _NETBSD_SOURCE if none of _ANSI_SOURCE,
_POSIX_C_SOURCE and _XOPEN_SOURCE is defined, so as to preserve
existing behaviour.

This has two major advantages:
+ Programs that require non-POSIX facilities but define _POSIX_C_SOURCE
can trivially be overruled by putting -D_NETBSD_SOURCE in their CFLAGS.
+ It makes most of the #ifs simpler, in that they're all now ORs of the
various macros, rather than having checks for (!defined(_ANSI_SOURCE) ||
!defined(_POSIX_C_SOURCE) || !defined(_XOPEN_SOURCE)) all over the place.

I've tried not to change the semantics of the headers in any case where
_NETBSD_SOURCE wasn't defined, but there were some places where the
current semantics were clearly mad, and retaining them was harder than
correcting them. In particular, I've mostly normalised things so that
_ANSI_SOURCE gets you the smallest set of stuff, then _POSIX_C_SOURCE,
_XOPEN_SOURCE and _NETBSD_SOURCE in that order.

Tested by building for vax, encouraged by thorpej, and uncontested in
tech-userlevel for a week.

Clean up some rampant code duplication wrt ieee number handling:
- Add alignment-safe double and float unions.
- Use the above for the __infinity and __nan constants on all
architectures that use the standard ieee754 representation of
those constants.
- Add a single copy of various ieee754 math functions (frexp, isinf,
isnan, ldexp and modf) that had numerous duplicates among the
arch-specific directories.
- Use the above functions on all architectures where the generic C
versions where used. Architectures that had local assembly
routines are untouched (for those functions only).

branches: 1.23.2;
protect the incomplete cabs declarations with #ifndef __MATH_PRIVATE__ so that
the code that defines the functions can declare them properly.

Remove 'extern' from function declarations.

branches: 1.21.2;
const -> __const and include <sys/cdefs.h> earlier; fixes PR lib/9052
by Takahiro Kambe.

C99: Define a NAN macro in <math.h> which evaulates to a constant expression of
a single-precision quiet NaN; only to be defined on platforms that do support
this value.

Add missing isinff() declaration.

This file has still missing prototypes: cabs and cabsf... At least comment
what the arguments should be.

s/void/const void/

cast to void * first to avoid lint warning.

Oops, isinf() was never included in any X/Open interface specification.

const poisoning.

Fix comment typo in previous.

Reorganize name space protection.

C++ has a very different concept of "exception." Delimit struct exception
and matherr() with #ifndef __cplusplus.

new RCS ID format.

Added prototypes for float versions of math functions.
Removed prototypes for internal functions.

Add cabs() and drem(), for BSD libm compatibility.

Add missing _. reported by several folks.

New math.h, based on fdlibm's fdlibm.h.

Moved definition of M_FOO constants into !_ANSI_SOURCE && !_POSIX_SOURCE
conditional.

Use __pure qualifier.

update for better FP routines, from AT&T & elsewhere

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision

Fix the rest of the functions that have been made visible by xopen5,
posix 2017, and c99.

PR/53393: coypu: move rint to the right section.

branches: 1.63.24;
Allow archs to provide __isinf() and __isnan() as inline functions instead
of macros.

If __FLT_EVAL_METHOD__ exists and is in {0,1,2}, provide float_t and
double_t.

FP_ILOGBNAN should be INT_MAX

Define FP_ILOGB0 and FP_ILOGBNAN

Add missing long double functions

branches: 1.58.2;
oops, forgot to commit: add tgamma and tgammaf

branches: 1.57.2;
Use __INFINITY in preference to __builtin_inff(). Currently only VAX defines
__INFINITY and allows to use __builtin_huge_valf() instead since gcc4.5 will
error out unconditionally if __builtin_inf() is used.

branches: 1.56.2;
Add ilogbl(3).

Add scalbnl to libm.

remquo{,f} from FreeBSD via Stathis Kamperis

branches: 1.53.2;
Commit SoC long double support from Stathis Kamperis

typo in comment

PR 36988: ISO C99 mandates that HUGE_VAL, HUGE_VALF, HUGE_VALL and
INFINITY are constant expressions.

Use the GCC builtins if possible to implement them.

branches: 1.50.2;
Add exp2 and exp2m

oops forgot to commit that one--- Add f{dim,min,max}{,l,f}

Make NAN a compile time constant (with help from the compiler). Suggested
by krister. Fixes PR 40695.
Make references to the old construct (which we can't remove for binary
compatibility reasons) emit a linker warning.

branches: 1.47.8; 1.47.10; 1.47.12; 1.47.16;
make name const.

branches: 1.46.10;
-get cabs() and cabsf() out of public view - these are historical and
conflict with C99 functions which are builtins in newer gcc
(actually, the old cabs() is ABI compatible with the new _complex one
on i386, but this is purely accidental)
remove public prototypes and manpages, move the code into a compat
subdirectory as libc does so that binary compatibility is kept
-add a manpage for the isgreater() etc macros, borrowed from FreeBSD

add the isgreater() at al. comparision macros defined in C99 and SUSv3;
use just the primitive macros for now (identical to FreeBSD/DragonFly)
which don't use gcc internals, the rest can go in after some testing;
addresses PR standards/25520

Add trunc() and truncf() C99 functions, from FreeBSD.

Add nan(3), nanf(3), and nanl(3); welcome libm.so.0.4.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

Add log2f

de-__P -- the hack is long since useless. Discussed with christos,
matt, kleink, others. Approved by christos.

Move round() to the C99 section.

lib/24252: Add C99 functions round(3) and roundf(3).
libm minor unchanged; ride a recent bump.

From FreeBSD.

being here, add the (l)lround(f) functions to libm
(almost the same like lrint)

shut up lint's "long long" warnings

Add (l)lrint(f).
Being here, move some C99 function declarations into the appropriate
section (as discussed with kleink).

branches: 1.34.2;
* Turn isinf(3) and isnan(3) into C99-style macros.
* Make it possible for ports to override these (i.e., VAX).
* Remove isnanl(), which was internal to libc only.

Make lint happy for __fpmacro_unary_floating() users.

Move FP_ROP and FP_DIRTYZERO from <math.h> to <vax/math.h> (where they
would have ended up if I hadn't missed the latter file in the commit).

Add FP_ROP and FP_DIRTYZERO to MD range for VAX.

Close the comment after #endif.

Add C99 fpclassify(), isfinite(), isnormal(), and signbit() macros.

C99 7.12#4: provide INFINITY, which defaults to HUGE_VALF.
If infinities are not available, the machine-dependent header must define
__INFINITY as a positive constant of type float that overflows.

C99: provide HUGE_VALF and HUGE_VALL.

Add long double support.

Add a new feature-test macro, _NETBSD_SOURCE. If this is defined
by the application, all NetBSD interfaces are made visible, even
if some other feature-test macro (like _POSIX_C_SOURCE) is defined.
<sys/featuretest.h> defined _NETBSD_SOURCE if none of _ANSI_SOURCE,
_POSIX_C_SOURCE and _XOPEN_SOURCE is defined, so as to preserve
existing behaviour.

This has two major advantages:
+ Programs that require non-POSIX facilities but define _POSIX_C_SOURCE
can trivially be overruled by putting -D_NETBSD_SOURCE in their CFLAGS.
+ It makes most of the #ifs simpler, in that they're all now ORs of the
various macros, rather than having checks for (!defined(_ANSI_SOURCE) ||
!defined(_POSIX_C_SOURCE) || !defined(_XOPEN_SOURCE)) all over the place.

I've tried not to change the semantics of the headers in any case where
_NETBSD_SOURCE wasn't defined, but there were some places where the
current semantics were clearly mad, and retaining them was harder than
correcting them. In particular, I've mostly normalised things so that
_ANSI_SOURCE gets you the smallest set of stuff, then _POSIX_C_SOURCE,
_XOPEN_SOURCE and _NETBSD_SOURCE in that order.

Tested by building for vax, encouraged by thorpej, and uncontested in
tech-userlevel for a week.

Clean up some rampant code duplication wrt ieee number handling:
- Add alignment-safe double and float unions.
- Use the above for the __infinity and __nan constants on all
architectures that use the standard ieee754 representation of
those constants.
- Add a single copy of various ieee754 math functions (frexp, isinf,
isnan, ldexp and modf) that had numerous duplicates among the
arch-specific directories.
- Use the above functions on all architectures where the generic C
versions where used. Architectures that had local assembly
routines are untouched (for those functions only).

branches: 1.23.2;
protect the incomplete cabs declarations with #ifndef __MATH_PRIVATE__ so that
the code that defines the functions can declare them properly.

Remove 'extern' from function declarations.

branches: 1.21.2;
const -> __const and include <sys/cdefs.h> earlier; fixes PR lib/9052
by Takahiro Kambe.

C99: Define a NAN macro in <math.h> which evaulates to a constant expression of
a single-precision quiet NaN; only to be defined on platforms that do support
this value.

Add missing isinff() declaration.

This file has still missing prototypes: cabs and cabsf... At least comment
what the arguments should be.

s/void/const void/

cast to void * first to avoid lint warning.

Oops, isinf() was never included in any X/Open interface specification.

const poisoning.

Fix comment typo in previous.

Reorganize name space protection.

C++ has a very different concept of "exception." Delimit struct exception
and matherr() with #ifndef __cplusplus.

new RCS ID format.

Added prototypes for float versions of math functions.
Removed prototypes for internal functions.

Add cabs() and drem(), for BSD libm compatibility.

Add missing _. reported by several folks.

New math.h, based on fdlibm's fdlibm.h.

Moved definition of M_FOO constants into !_ANSI_SOURCE && !_POSIX_SOURCE
conditional.

Use __pure qualifier.

update for better FP routines, from AT&T & elsewhere

after 0.2.2 "stable" patches applied

branches: 1.1.1;
Initial revision