put reallocarray() in the public namespace

reallocarray() will be part of the next POSIX release, see
https://austingroupbugs.net/view.php?id=1218

adapt an errno value to match POSIX expectations

As discussed on tech-userlevel

Remove hacky gcc2-specific alloca definition.

PR/56250: Michael Forney: mkstemp/mkdtemp not declared with
_POSIX_C_SOURCE=200809L

Avoid depending on namespace pollution in stdlib.h.

We can make a stronger guarantee than __pure.
abs/labs don't read global memory.

Make GCC 9 happier.

PR standards/53525: move getsubopt() to stdlib.h

branches: 1.119.2; 1.119.4;
remove recallocarray

add recallocarray (used by openssh-7.6)

add hmac(3)

Add the C11 function aligned_alloc to libc.

Move reallocarray under _OPENBSD_SOURCE where it belongs.

Introduce reallocarr(3) for easy (re)allocation of memory with overflow
checks. Compared to OpenBSD's reallocarray, makes it easier to avoid memory
leaks on allocation failures and it doesn't depend on malloc(0) != NULL
for correct error checking. Compared to plain realloc, it also avoids
the problem of intermediate integer overflows. The trade-off is the use
of void * to side step C type system with regard to generic pointer to
pointer.

add reallocarray

remove restrict to match with OpenBSD prototype (Kamil Rytarowski)

Add strtonum protected with _OPENBSD_SOURCE

add mkostemp{s,} mkstemps from FreeBSD

rename to 60

fix srandom and initstate signatures (from enh at google)

branches: 1.107.2;
add ptsname_r

add at_quick_exit and quick_exit

Provide the long long routines for C++11 code.

Add MB_CUR_MAX_L.

Add mblen_l, mbstowcs_l, wctomb_l, mbtowc_l and wcstombs_l.

Add strtof_l, strtod_l and strtold_l.

Add strtol_l and friends. Switch _citrus_bcs_strtol to use plain
strtol_l unless in tools mode. Add note to retire the BCS code on the
next libc major bump.

branches: 1.100.2;
sprinkle restrict in realpath(3)

Switch device database to cdb(5). Rework ttyname(3) and ttyname_r(3) to
depend on new devname_r(3) as heart. Add /dev/pts magic directly to
devname(3). While it can lead to returning non-existing paths, the
behavior is more consistent that way. Drop caching layer in devname(3),
it doesn't buy anything for the common case of having access to the
database. Teach devname(3) proper fallback behavior of scanning /dev.
Create both old-style and new-style database for now in /etc/rc.d/sysdb.

PR/46360: YAMAMOTO Takashi: Restore NetBSD-5 compatibility with putenv()
copying the passed string (which is not ToG compliant), instead of using
it directly in the environment arrat as it should. Needs to be pulled up
to NetBSd-6.

branches: 1.97.4; 1.97.6;
remove commaize_number.

put in a proto for commaize_number()
(nb. doing this purely as the minimum work solution to get a working build)

add arc4random_{buf,uniform}, from OpenBSD.

branches: 1.94.2;
Use __constfunc(3) instead of __attribute__((const)) directly.

make putenv POSIX compliant.

later versions of pcc may define __GNUC__ for compatibility reasons, so
having a system definition for alloca is counter productive as it seems
fairly common to use

#ifdef __GNUC__
#define alloca ..
#endif

in user-headers (eg see usr.sbin/gspa/gspa/gspa_ass.h)

So, defang this definition as we don't otherwise provide alloca on NetBSD

Consistently use __printflike and __scanflike.

branches: 1.90.2;
Don't define wchar_t for c++; it is native in the language. From Alexander
Nasonov

Add a fast, platform independent hash function to libc.
The algorithm used is the Jenkins hash. The name (mi_vector_hash)
reflects the nature of the hash function.
Add glue for libc ATF tests and include a test case to make sure that
(mis)alignment and endianess are handled correctly.

Bump libc minor to 169.

cleanup after devmajor_t:
-since getdevmajor(3) is now binary compatible again with <=5.0
there is no need to rename, I've just left a __getdevmajor50 symbol
temporarily for those who track -current
-update manpage

Change major()/minor() to return 32-bit types again, called
devmajor_t/devminor_t, as proposed on tech-kern.
This avoids 64-bit arithmetics and 64-bit printf formats in parts
of the kernel where it is not really useful, and helps clarity.

merge christos-time_t

branches: 1.85.4; 1.85.6; 1.85.12;
Use __builtin_alloc() with pcc.

branches: 1.84.4; 1.84.6;
RANDOM_MAX: don't rely on integer overflow.

everyone else assumes that defining "_XOPEN_SOURCE" means XPG4.2,
so remove comparisions against the numerical values "3" or "4"

add RANDOM_MAX define from anon ymous

branches: 1.81.2;
Add csetexpandtc()

change some __attribute__s to __keywords

humanize_number was the only function in <util.h> that is not in libutil.
Move it to stdlib.h and add dehumanize_number.

Enable posix_memalign().

branches: 1.77.4;
The alloca() arg type check is broken (cpp do not do string comparisions)
so make it marginally less broken so that it works with other compilers
than gcc.
Probably the check can be removed, I doubt anyone will try to use gcc1
anymore.

Add strtof(3) and strtold(3); welcome libc 12.137.

u_intN_t -> uintN_t

__const__ -> const
This is in an __attribute__ so it isn't really necessary, but it gets
it out of the listing of __keyword__s...

Move getenv_r() to the local section.

Add getenv_r and ttyname_r

compat code reorg

Add a macro __MKTEMP_OK__ that renames mktemp() to _mktemp() for uses that
have been checked, so that the linker does not warn us. There are valid
uses for mktemp() such as for creating filenames that are going to be
targets of the rename system call. Discussed with lukem.

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

Add a function called getdevmajor().

It tells you the major device number for whatever character or block
device you ask it. This is sort of the inverse of devname(3) but not
quite, since it's backed by the kernel (sysctl's kern.drivers
information) and not a database cobbled together from the contents of
the filesystem.

Add unix/98 pty function prototypes.

more const sprinkling on cget* suggested by kleink.

constify the db name vector argument.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22270, verified by myself.

Do not give names to the parameters in function prototypes (those names
reduces the usefulness of the gcc -Wshadow flag).

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.

Well, _The_ Open Group.

Use the real TOG specification name for references.

1003.1-2001 has introduced unsetenv() which differs from the current
implementation in not permitting a "name=value" argument.
* Add a conforming __unsetenv13() and do function renaming for
unsetenv(); preserve old symbol with old behavior.
* Make visible setenv() and unsetenv() for 1003.1-2001 feature selection
macros; resolves PR standards/20479.

Add _Exit(), for C99 and POSIX-2001.
Proposed on tech-userlevel to a resounding silence in April last year.

add strsuftoll(3) and strsuftollx(3) - parse a string for a number
with an optional suffix, product (two numbers separated by `x'), and
provide range checking

add arc4random(3). from openbsd

branches: 1.55.2;
Don't declare qdiv() when qdiv_t wasn't defined previously; fixes PR lib/11557
from Richard Earnshaw.

Sprinkle some restrict qualifiers.

add prototypes for shquote() and shquotev()

branches: 1.52.2;
add prototypes for getprogname and setprogname

deal with bsearch duplicate declaration.

Move fmtcheck() prototype to stdio.h. It makes more sense to declare it
with the formatting functions with which it is meant to be used. Thanks
to Klaus Klein for "encouraging" me to make this change.

Add {__,}fmtcheck(), a function specified by Bill Sommerfeld to check
a user-supplied format string against a fallback format string to ensure
that they will consume arguments of the same type.

libc minor bump for new API entry point.

Allow format-arg checking for setproctitle

ISO/IEC 9899:1999: MB_CUR_MAX expands to a positive integer expression with
type size_t (not int).

(Incompatible ABI change, ignored due to a window of just two days.)

Preparation for the future introduction of multibyte locale.
- MB_LEN_MAX is increased to 32.
- To ensure binary compatibility for old executables
under multibyte locale, versioned setlocale is added.
- __mb_len_cur definision is added in setlocale.c
and enable it in stdlib.h .
It is also important for multibyte locale stuffs,
but I just forgot.

branches: 1.45.2; 1.45.4;
Add ISO C99 long long integer general utility interfaces; partially addresses
PR standards/9482.

Move <null.h> to <sys/null.h> for the purpose of avoiding conflicts due to
std C headers including it, and a kernel module written in C++ would benfit
as well.

For the benefit of G++'s null pointer constant implementation, #define NULL
as __null with egcs 1.0 (GCC 2.90) and above. As several headers are affected
by this change, move the definition into a new header file, <null.h>, to ease
maintenance.

Add l64a_r(), a reentrant version of l64a(); found in the Multithreading
Extension of SVID4.

__Namespace__ __protection__.

Sprinkle in some __attribute__((noreturn))s.

Actually, FreeBSD calls this mkdtemp(). No point in creating a new name.

Add a mktempdir(), like mkstemp(), but it creates a directory.

Slight rearrangement.

If __AUDIT__ is defined, remove some traditional interfaces: gets, sprintf,
vsprintf, tempnam, mktemp, getwd.

Mark more functions returning internal buffers for const auditing.

const poisoning.

* Further name space protection reorganization: distinguish between various
issues of the XPG.
* Move setkey() prototype from <unistd.h> to this file. (XPG4)
* Move mkstemp(), mktemp(), ttyslot() and valloc() prototypes from <unistd.h>
to this file. (XPG4.2)

Reorganize name space protection.

Per XPG4.2, change the `n' argument of initstate() to size_t. (Since it was
recently changed from int to long (lite-2 import), I saw no reason to delay
this `alignment'.)

merge lite-2

Fix the prototypes for srandom() and initstate() after the libc/stdlib lite-2
merge (ints -> longs). The only platform on which this might have caused
a problem is the Alpha, and it happens to be safe because the first 6 arguments
are passed in registers on the Alpha (and are thus implicitly promoted to
longs anyhow).

devname is (dev_t, mode_t) not (int, int)

some minor KNF

Add missing rand_r prototype

Add __dead qualifier to abort() and exit(); PR 1845

added prototype for qabs()

Moved conditional #include of <sys/types.h> to near the top of file.
Defined qdiv_t typedef.
Declared qdiv function.

fix typo

add prototype for realpath()

new RCS ID format.

Add strtoq() & strtouq() declarations

add 3 protos

Update *sort() prototypes.

add _BSD's to ansi types

add new getopt(3) extern variable optreset

setproctitle proto

kill third arg to getbsize.

getbsize takes a (somewhat bogus, IMO) third parameter

add protos for getbsize() (in libutil) and the getcap routines (in libc)

stdlib.h should define NULL -- Ansi C, Section 7.10.

Fix typo in last checkin.

Added prototypes for a64l() & l64a().

Moved declarations for cfree, setenv, putenv and the rand48 family
into !_ANSI_SOURCE && !_POSIX_SOURCE conditional.
Added extern for getopt's optopt variable.

#include <machine/ansi.h>, declare size_t typedef.
Only include <sys/types.h> if !_POSIX_SOURCE.

Rand48 routines from martin@ntkupc1.tuwien.ac.at (Martin Birgmeier)

Add RCS identifiers.

GCC2 defines alloca() differently.

patch from rich murphy (rich@rice.edu) to fix alloca prototype

branches: 1.1.1;
Initial revision

Remove hacky gcc2-specific alloca definition.

PR/56250: Michael Forney: mkstemp/mkdtemp not declared with
_POSIX_C_SOURCE=200809L

Avoid depending on namespace pollution in stdlib.h.

We can make a stronger guarantee than __pure.
abs/labs don't read global memory.

Make GCC 9 happier.

PR standards/53525: move getsubopt() to stdlib.h

branches: 1.119.2; 1.119.4;
remove recallocarray

add recallocarray (used by openssh-7.6)

add hmac(3)

Add the C11 function aligned_alloc to libc.

Move reallocarray under _OPENBSD_SOURCE where it belongs.

Introduce reallocarr(3) for easy (re)allocation of memory with overflow
checks. Compared to OpenBSD's reallocarray, makes it easier to avoid memory
leaks on allocation failures and it doesn't depend on malloc(0) != NULL
for correct error checking. Compared to plain realloc, it also avoids
the problem of intermediate integer overflows. The trade-off is the use
of void * to side step C type system with regard to generic pointer to
pointer.

add reallocarray

remove restrict to match with OpenBSD prototype (Kamil Rytarowski)

Add strtonum protected with _OPENBSD_SOURCE

add mkostemp{s,} mkstemps from FreeBSD

rename to 60

fix srandom and initstate signatures (from enh at google)

branches: 1.107.2;
add ptsname_r

add at_quick_exit and quick_exit

Provide the long long routines for C++11 code.

Add MB_CUR_MAX_L.

Add mblen_l, mbstowcs_l, wctomb_l, mbtowc_l and wcstombs_l.

Add strtof_l, strtod_l and strtold_l.

Add strtol_l and friends. Switch _citrus_bcs_strtol to use plain
strtol_l unless in tools mode. Add note to retire the BCS code on the
next libc major bump.

branches: 1.100.2;
sprinkle restrict in realpath(3)

Switch device database to cdb(5). Rework ttyname(3) and ttyname_r(3) to
depend on new devname_r(3) as heart. Add /dev/pts magic directly to
devname(3). While it can lead to returning non-existing paths, the
behavior is more consistent that way. Drop caching layer in devname(3),
it doesn't buy anything for the common case of having access to the
database. Teach devname(3) proper fallback behavior of scanning /dev.
Create both old-style and new-style database for now in /etc/rc.d/sysdb.

PR/46360: YAMAMOTO Takashi: Restore NetBSD-5 compatibility with putenv()
copying the passed string (which is not ToG compliant), instead of using
it directly in the environment arrat as it should. Needs to be pulled up
to NetBSd-6.

branches: 1.97.4; 1.97.6;
remove commaize_number.

put in a proto for commaize_number()
(nb. doing this purely as the minimum work solution to get a working build)

add arc4random_{buf,uniform}, from OpenBSD.

branches: 1.94.2;
Use __constfunc(3) instead of __attribute__((const)) directly.

make putenv POSIX compliant.

later versions of pcc may define __GNUC__ for compatibility reasons, so
having a system definition for alloca is counter productive as it seems
fairly common to use

#ifdef __GNUC__
#define alloca ..
#endif

in user-headers (eg see usr.sbin/gspa/gspa/gspa_ass.h)

So, defang this definition as we don't otherwise provide alloca on NetBSD

Consistently use __printflike and __scanflike.

branches: 1.90.2;
Don't define wchar_t for c++; it is native in the language. From Alexander
Nasonov

Add a fast, platform independent hash function to libc.
The algorithm used is the Jenkins hash. The name (mi_vector_hash)
reflects the nature of the hash function.
Add glue for libc ATF tests and include a test case to make sure that
(mis)alignment and endianess are handled correctly.

Bump libc minor to 169.

cleanup after devmajor_t:
-since getdevmajor(3) is now binary compatible again with <=5.0
there is no need to rename, I've just left a __getdevmajor50 symbol
temporarily for those who track -current
-update manpage

Change major()/minor() to return 32-bit types again, called
devmajor_t/devminor_t, as proposed on tech-kern.
This avoids 64-bit arithmetics and 64-bit printf formats in parts
of the kernel where it is not really useful, and helps clarity.

merge christos-time_t

branches: 1.85.4; 1.85.6; 1.85.12;
Use __builtin_alloc() with pcc.

branches: 1.84.4; 1.84.6;
RANDOM_MAX: don't rely on integer overflow.

everyone else assumes that defining "_XOPEN_SOURCE" means XPG4.2,
so remove comparisions against the numerical values "3" or "4"

add RANDOM_MAX define from anon ymous

branches: 1.81.2;
Add csetexpandtc()

change some __attribute__s to __keywords

humanize_number was the only function in <util.h> that is not in libutil.
Move it to stdlib.h and add dehumanize_number.

Enable posix_memalign().

branches: 1.77.4;
The alloca() arg type check is broken (cpp do not do string comparisions)
so make it marginally less broken so that it works with other compilers
than gcc.
Probably the check can be removed, I doubt anyone will try to use gcc1
anymore.

Add strtof(3) and strtold(3); welcome libc 12.137.

u_intN_t -> uintN_t

__const__ -> const
This is in an __attribute__ so it isn't really necessary, but it gets
it out of the listing of __keyword__s...

Move getenv_r() to the local section.

Add getenv_r and ttyname_r

compat code reorg

Add a macro __MKTEMP_OK__ that renames mktemp() to _mktemp() for uses that
have been checked, so that the linker does not warn us. There are valid
uses for mktemp() such as for creating filenames that are going to be
targets of the rename system call. Discussed with lukem.

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

Add a function called getdevmajor().

It tells you the major device number for whatever character or block
device you ask it. This is sort of the inverse of devname(3) but not
quite, since it's backed by the kernel (sysctl's kern.drivers
information) and not a database cobbled together from the contents of
the filesystem.

Add unix/98 pty function prototypes.

more const sprinkling on cget* suggested by kleink.

constify the db name vector argument.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22270, verified by myself.

Do not give names to the parameters in function prototypes (those names
reduces the usefulness of the gcc -Wshadow flag).

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.

Well, _The_ Open Group.

Use the real TOG specification name for references.

1003.1-2001 has introduced unsetenv() which differs from the current
implementation in not permitting a "name=value" argument.
* Add a conforming __unsetenv13() and do function renaming for
unsetenv(); preserve old symbol with old behavior.
* Make visible setenv() and unsetenv() for 1003.1-2001 feature selection
macros; resolves PR standards/20479.

Add _Exit(), for C99 and POSIX-2001.
Proposed on tech-userlevel to a resounding silence in April last year.

add strsuftoll(3) and strsuftollx(3) - parse a string for a number
with an optional suffix, product (two numbers separated by `x'), and
provide range checking

add arc4random(3). from openbsd

branches: 1.55.2;
Don't declare qdiv() when qdiv_t wasn't defined previously; fixes PR lib/11557
from Richard Earnshaw.

Sprinkle some restrict qualifiers.

add prototypes for shquote() and shquotev()

branches: 1.52.2;
add prototypes for getprogname and setprogname

deal with bsearch duplicate declaration.

Move fmtcheck() prototype to stdio.h. It makes more sense to declare it
with the formatting functions with which it is meant to be used. Thanks
to Klaus Klein for "encouraging" me to make this change.

Add {__,}fmtcheck(), a function specified by Bill Sommerfeld to check
a user-supplied format string against a fallback format string to ensure
that they will consume arguments of the same type.

libc minor bump for new API entry point.

Allow format-arg checking for setproctitle

ISO/IEC 9899:1999: MB_CUR_MAX expands to a positive integer expression with
type size_t (not int).

(Incompatible ABI change, ignored due to a window of just two days.)

Preparation for the future introduction of multibyte locale.
- MB_LEN_MAX is increased to 32.
- To ensure binary compatibility for old executables
under multibyte locale, versioned setlocale is added.
- __mb_len_cur definision is added in setlocale.c
and enable it in stdlib.h .
It is also important for multibyte locale stuffs,
but I just forgot.

branches: 1.45.2; 1.45.4;
Add ISO C99 long long integer general utility interfaces; partially addresses
PR standards/9482.

Move <null.h> to <sys/null.h> for the purpose of avoiding conflicts due to
std C headers including it, and a kernel module written in C++ would benfit
as well.

For the benefit of G++'s null pointer constant implementation, #define NULL
as __null with egcs 1.0 (GCC 2.90) and above. As several headers are affected
by this change, move the definition into a new header file, <null.h>, to ease
maintenance.

Add l64a_r(), a reentrant version of l64a(); found in the Multithreading
Extension of SVID4.

__Namespace__ __protection__.

Sprinkle in some __attribute__((noreturn))s.

Actually, FreeBSD calls this mkdtemp(). No point in creating a new name.

Add a mktempdir(), like mkstemp(), but it creates a directory.

Slight rearrangement.

If __AUDIT__ is defined, remove some traditional interfaces: gets, sprintf,
vsprintf, tempnam, mktemp, getwd.

Mark more functions returning internal buffers for const auditing.

const poisoning.

* Further name space protection reorganization: distinguish between various
issues of the XPG.
* Move setkey() prototype from <unistd.h> to this file. (XPG4)
* Move mkstemp(), mktemp(), ttyslot() and valloc() prototypes from <unistd.h>
to this file. (XPG4.2)

Reorganize name space protection.

Per XPG4.2, change the `n' argument of initstate() to size_t. (Since it was
recently changed from int to long (lite-2 import), I saw no reason to delay
this `alignment'.)

merge lite-2

Fix the prototypes for srandom() and initstate() after the libc/stdlib lite-2
merge (ints -> longs). The only platform on which this might have caused
a problem is the Alpha, and it happens to be safe because the first 6 arguments
are passed in registers on the Alpha (and are thus implicitly promoted to
longs anyhow).

devname is (dev_t, mode_t) not (int, int)

some minor KNF

Add missing rand_r prototype

Add __dead qualifier to abort() and exit(); PR 1845

added prototype for qabs()

Moved conditional #include of <sys/types.h> to near the top of file.
Defined qdiv_t typedef.
Declared qdiv function.

fix typo

add prototype for realpath()

new RCS ID format.

Add strtoq() & strtouq() declarations

add 3 protos

Update *sort() prototypes.

add _BSD's to ansi types

add new getopt(3) extern variable optreset

setproctitle proto

kill third arg to getbsize.

getbsize takes a (somewhat bogus, IMO) third parameter

add protos for getbsize() (in libutil) and the getcap routines (in libc)

stdlib.h should define NULL -- Ansi C, Section 7.10.

Fix typo in last checkin.

Added prototypes for a64l() & l64a().

Moved declarations for cfree, setenv, putenv and the rand48 family
into !_ANSI_SOURCE && !_POSIX_SOURCE conditional.
Added extern for getopt's optopt variable.

#include <machine/ansi.h>, declare size_t typedef.
Only include <sys/types.h> if !_POSIX_SOURCE.

Rand48 routines from martin@ntkupc1.tuwien.ac.at (Martin Birgmeier)

Add RCS identifiers.

GCC2 defines alloca() differently.

patch from rich murphy (rich@rice.edu) to fix alloca prototype

branches: 1.1.1;
Initial revision

PR/56250: Michael Forney: mkstemp/mkdtemp not declared with
_POSIX_C_SOURCE=200809L

Avoid depending on namespace pollution in stdlib.h.

We can make a stronger guarantee than __pure.
abs/labs don't read global memory.

Make GCC 9 happier.

PR standards/53525: move getsubopt() to stdlib.h

branches: 1.119.2; 1.119.4;
remove recallocarray

add recallocarray (used by openssh-7.6)

add hmac(3)

Add the C11 function aligned_alloc to libc.

Move reallocarray under _OPENBSD_SOURCE where it belongs.

Introduce reallocarr(3) for easy (re)allocation of memory with overflow
checks. Compared to OpenBSD's reallocarray, makes it easier to avoid memory
leaks on allocation failures and it doesn't depend on malloc(0) != NULL
for correct error checking. Compared to plain realloc, it also avoids
the problem of intermediate integer overflows. The trade-off is the use
of void * to side step C type system with regard to generic pointer to
pointer.

add reallocarray

remove restrict to match with OpenBSD prototype (Kamil Rytarowski)

Add strtonum protected with _OPENBSD_SOURCE

add mkostemp{s,} mkstemps from FreeBSD

rename to 60

fix srandom and initstate signatures (from enh at google)

branches: 1.107.2;
add ptsname_r

add at_quick_exit and quick_exit

Provide the long long routines for C++11 code.

Add MB_CUR_MAX_L.

Add mblen_l, mbstowcs_l, wctomb_l, mbtowc_l and wcstombs_l.

Add strtof_l, strtod_l and strtold_l.

Add strtol_l and friends. Switch _citrus_bcs_strtol to use plain
strtol_l unless in tools mode. Add note to retire the BCS code on the
next libc major bump.

branches: 1.100.2;
sprinkle restrict in realpath(3)

Switch device database to cdb(5). Rework ttyname(3) and ttyname_r(3) to
depend on new devname_r(3) as heart. Add /dev/pts magic directly to
devname(3). While it can lead to returning non-existing paths, the
behavior is more consistent that way. Drop caching layer in devname(3),
it doesn't buy anything for the common case of having access to the
database. Teach devname(3) proper fallback behavior of scanning /dev.
Create both old-style and new-style database for now in /etc/rc.d/sysdb.

PR/46360: YAMAMOTO Takashi: Restore NetBSD-5 compatibility with putenv()
copying the passed string (which is not ToG compliant), instead of using
it directly in the environment arrat as it should. Needs to be pulled up
to NetBSd-6.

branches: 1.97.4; 1.97.6;
remove commaize_number.

put in a proto for commaize_number()
(nb. doing this purely as the minimum work solution to get a working build)

add arc4random_{buf,uniform}, from OpenBSD.

branches: 1.94.2;
Use __constfunc(3) instead of __attribute__((const)) directly.

make putenv POSIX compliant.

later versions of pcc may define __GNUC__ for compatibility reasons, so
having a system definition for alloca is counter productive as it seems
fairly common to use

#ifdef __GNUC__
#define alloca ..
#endif

in user-headers (eg see usr.sbin/gspa/gspa/gspa_ass.h)

So, defang this definition as we don't otherwise provide alloca on NetBSD

Consistently use __printflike and __scanflike.

branches: 1.90.2;
Don't define wchar_t for c++; it is native in the language. From Alexander
Nasonov

Add a fast, platform independent hash function to libc.
The algorithm used is the Jenkins hash. The name (mi_vector_hash)
reflects the nature of the hash function.
Add glue for libc ATF tests and include a test case to make sure that
(mis)alignment and endianess are handled correctly.

Bump libc minor to 169.

cleanup after devmajor_t:
-since getdevmajor(3) is now binary compatible again with <=5.0
there is no need to rename, I've just left a __getdevmajor50 symbol
temporarily for those who track -current
-update manpage

Change major()/minor() to return 32-bit types again, called
devmajor_t/devminor_t, as proposed on tech-kern.
This avoids 64-bit arithmetics and 64-bit printf formats in parts
of the kernel where it is not really useful, and helps clarity.

merge christos-time_t

branches: 1.85.4; 1.85.6; 1.85.12;
Use __builtin_alloc() with pcc.

branches: 1.84.4; 1.84.6;
RANDOM_MAX: don't rely on integer overflow.

everyone else assumes that defining "_XOPEN_SOURCE" means XPG4.2,
so remove comparisions against the numerical values "3" or "4"

add RANDOM_MAX define from anon ymous

branches: 1.81.2;
Add csetexpandtc()

change some __attribute__s to __keywords

humanize_number was the only function in <util.h> that is not in libutil.
Move it to stdlib.h and add dehumanize_number.

Enable posix_memalign().

branches: 1.77.4;
The alloca() arg type check is broken (cpp do not do string comparisions)
so make it marginally less broken so that it works with other compilers
than gcc.
Probably the check can be removed, I doubt anyone will try to use gcc1
anymore.

Add strtof(3) and strtold(3); welcome libc 12.137.

u_intN_t -> uintN_t

__const__ -> const
This is in an __attribute__ so it isn't really necessary, but it gets
it out of the listing of __keyword__s...

Move getenv_r() to the local section.

Add getenv_r and ttyname_r

compat code reorg

Add a macro __MKTEMP_OK__ that renames mktemp() to _mktemp() for uses that
have been checked, so that the linker does not warn us. There are valid
uses for mktemp() such as for creating filenames that are going to be
targets of the rename system call. Discussed with lukem.

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

Add a function called getdevmajor().

It tells you the major device number for whatever character or block
device you ask it. This is sort of the inverse of devname(3) but not
quite, since it's backed by the kernel (sysctl's kern.drivers
information) and not a database cobbled together from the contents of
the filesystem.

Add unix/98 pty function prototypes.

more const sprinkling on cget* suggested by kleink.

constify the db name vector argument.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22270, verified by myself.

Do not give names to the parameters in function prototypes (those names
reduces the usefulness of the gcc -Wshadow flag).

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.

Well, _The_ Open Group.

Use the real TOG specification name for references.

1003.1-2001 has introduced unsetenv() which differs from the current
implementation in not permitting a "name=value" argument.
* Add a conforming __unsetenv13() and do function renaming for
unsetenv(); preserve old symbol with old behavior.
* Make visible setenv() and unsetenv() for 1003.1-2001 feature selection
macros; resolves PR standards/20479.

Add _Exit(), for C99 and POSIX-2001.
Proposed on tech-userlevel to a resounding silence in April last year.

add strsuftoll(3) and strsuftollx(3) - parse a string for a number
with an optional suffix, product (two numbers separated by `x'), and
provide range checking

add arc4random(3). from openbsd

branches: 1.55.2;
Don't declare qdiv() when qdiv_t wasn't defined previously; fixes PR lib/11557
from Richard Earnshaw.

Sprinkle some restrict qualifiers.

add prototypes for shquote() and shquotev()

branches: 1.52.2;
add prototypes for getprogname and setprogname

deal with bsearch duplicate declaration.

Move fmtcheck() prototype to stdio.h. It makes more sense to declare it
with the formatting functions with which it is meant to be used. Thanks
to Klaus Klein for "encouraging" me to make this change.

Add {__,}fmtcheck(), a function specified by Bill Sommerfeld to check
a user-supplied format string against a fallback format string to ensure
that they will consume arguments of the same type.

libc minor bump for new API entry point.

Allow format-arg checking for setproctitle

ISO/IEC 9899:1999: MB_CUR_MAX expands to a positive integer expression with
type size_t (not int).

(Incompatible ABI change, ignored due to a window of just two days.)

Preparation for the future introduction of multibyte locale.
- MB_LEN_MAX is increased to 32.
- To ensure binary compatibility for old executables
under multibyte locale, versioned setlocale is added.
- __mb_len_cur definision is added in setlocale.c
and enable it in stdlib.h .
It is also important for multibyte locale stuffs,
but I just forgot.

branches: 1.45.2; 1.45.4;
Add ISO C99 long long integer general utility interfaces; partially addresses
PR standards/9482.

Move <null.h> to <sys/null.h> for the purpose of avoiding conflicts due to
std C headers including it, and a kernel module written in C++ would benfit
as well.

For the benefit of G++'s null pointer constant implementation, #define NULL
as __null with egcs 1.0 (GCC 2.90) and above. As several headers are affected
by this change, move the definition into a new header file, <null.h>, to ease
maintenance.

Add l64a_r(), a reentrant version of l64a(); found in the Multithreading
Extension of SVID4.

__Namespace__ __protection__.

Sprinkle in some __attribute__((noreturn))s.

Actually, FreeBSD calls this mkdtemp(). No point in creating a new name.

Add a mktempdir(), like mkstemp(), but it creates a directory.

Slight rearrangement.

If __AUDIT__ is defined, remove some traditional interfaces: gets, sprintf,
vsprintf, tempnam, mktemp, getwd.

Mark more functions returning internal buffers for const auditing.

const poisoning.

* Further name space protection reorganization: distinguish between various
issues of the XPG.
* Move setkey() prototype from <unistd.h> to this file. (XPG4)
* Move mkstemp(), mktemp(), ttyslot() and valloc() prototypes from <unistd.h>
to this file. (XPG4.2)

Reorganize name space protection.

Per XPG4.2, change the `n' argument of initstate() to size_t. (Since it was
recently changed from int to long (lite-2 import), I saw no reason to delay
this `alignment'.)

merge lite-2

Fix the prototypes for srandom() and initstate() after the libc/stdlib lite-2
merge (ints -> longs). The only platform on which this might have caused
a problem is the Alpha, and it happens to be safe because the first 6 arguments
are passed in registers on the Alpha (and are thus implicitly promoted to
longs anyhow).

devname is (dev_t, mode_t) not (int, int)

some minor KNF

Add missing rand_r prototype

Add __dead qualifier to abort() and exit(); PR 1845

added prototype for qabs()

Moved conditional #include of <sys/types.h> to near the top of file.
Defined qdiv_t typedef.
Declared qdiv function.

fix typo

add prototype for realpath()

new RCS ID format.

Add strtoq() & strtouq() declarations

add 3 protos

Update *sort() prototypes.

add _BSD's to ansi types

add new getopt(3) extern variable optreset

setproctitle proto

kill third arg to getbsize.

getbsize takes a (somewhat bogus, IMO) third parameter

add protos for getbsize() (in libutil) and the getcap routines (in libc)

stdlib.h should define NULL -- Ansi C, Section 7.10.

Fix typo in last checkin.

Added prototypes for a64l() & l64a().

Moved declarations for cfree, setenv, putenv and the rand48 family
into !_ANSI_SOURCE && !_POSIX_SOURCE conditional.
Added extern for getopt's optopt variable.

#include <machine/ansi.h>, declare size_t typedef.
Only include <sys/types.h> if !_POSIX_SOURCE.

Rand48 routines from martin@ntkupc1.tuwien.ac.at (Martin Birgmeier)

Add RCS identifiers.

GCC2 defines alloca() differently.

patch from rich murphy (rich@rice.edu) to fix alloca prototype

branches: 1.1.1;
Initial revision

Avoid depending on namespace pollution in stdlib.h.

We can make a stronger guarantee than __pure.
abs/labs don't read global memory.

Make GCC 9 happier.

PR standards/53525: move getsubopt() to stdlib.h

branches: 1.119.2; 1.119.4;
remove recallocarray

add recallocarray (used by openssh-7.6)

add hmac(3)

Add the C11 function aligned_alloc to libc.

Move reallocarray under _OPENBSD_SOURCE where it belongs.

Introduce reallocarr(3) for easy (re)allocation of memory with overflow
checks. Compared to OpenBSD's reallocarray, makes it easier to avoid memory
leaks on allocation failures and it doesn't depend on malloc(0) != NULL
for correct error checking. Compared to plain realloc, it also avoids
the problem of intermediate integer overflows. The trade-off is the use
of void * to side step C type system with regard to generic pointer to
pointer.

add reallocarray

remove restrict to match with OpenBSD prototype (Kamil Rytarowski)

Add strtonum protected with _OPENBSD_SOURCE

add mkostemp{s,} mkstemps from FreeBSD

rename to 60

fix srandom and initstate signatures (from enh at google)

branches: 1.107.2;
add ptsname_r

add at_quick_exit and quick_exit

Provide the long long routines for C++11 code.

Add MB_CUR_MAX_L.

Add mblen_l, mbstowcs_l, wctomb_l, mbtowc_l and wcstombs_l.

Add strtof_l, strtod_l and strtold_l.

Add strtol_l and friends. Switch _citrus_bcs_strtol to use plain
strtol_l unless in tools mode. Add note to retire the BCS code on the
next libc major bump.

branches: 1.100.2;
sprinkle restrict in realpath(3)

Switch device database to cdb(5). Rework ttyname(3) and ttyname_r(3) to
depend on new devname_r(3) as heart. Add /dev/pts magic directly to
devname(3). While it can lead to returning non-existing paths, the
behavior is more consistent that way. Drop caching layer in devname(3),
it doesn't buy anything for the common case of having access to the
database. Teach devname(3) proper fallback behavior of scanning /dev.
Create both old-style and new-style database for now in /etc/rc.d/sysdb.

PR/46360: YAMAMOTO Takashi: Restore NetBSD-5 compatibility with putenv()
copying the passed string (which is not ToG compliant), instead of using
it directly in the environment arrat as it should. Needs to be pulled up
to NetBSd-6.

branches: 1.97.4; 1.97.6;
remove commaize_number.

put in a proto for commaize_number()
(nb. doing this purely as the minimum work solution to get a working build)

add arc4random_{buf,uniform}, from OpenBSD.

branches: 1.94.2;
Use __constfunc(3) instead of __attribute__((const)) directly.

make putenv POSIX compliant.

later versions of pcc may define __GNUC__ for compatibility reasons, so
having a system definition for alloca is counter productive as it seems
fairly common to use

#ifdef __GNUC__
#define alloca ..
#endif

in user-headers (eg see usr.sbin/gspa/gspa/gspa_ass.h)

So, defang this definition as we don't otherwise provide alloca on NetBSD

Consistently use __printflike and __scanflike.

branches: 1.90.2;
Don't define wchar_t for c++; it is native in the language. From Alexander
Nasonov

Add a fast, platform independent hash function to libc.
The algorithm used is the Jenkins hash. The name (mi_vector_hash)
reflects the nature of the hash function.
Add glue for libc ATF tests and include a test case to make sure that
(mis)alignment and endianess are handled correctly.

Bump libc minor to 169.

cleanup after devmajor_t:
-since getdevmajor(3) is now binary compatible again with <=5.0
there is no need to rename, I've just left a __getdevmajor50 symbol
temporarily for those who track -current
-update manpage

Change major()/minor() to return 32-bit types again, called
devmajor_t/devminor_t, as proposed on tech-kern.
This avoids 64-bit arithmetics and 64-bit printf formats in parts
of the kernel where it is not really useful, and helps clarity.

merge christos-time_t

branches: 1.85.4; 1.85.6; 1.85.12;
Use __builtin_alloc() with pcc.

branches: 1.84.4; 1.84.6;
RANDOM_MAX: don't rely on integer overflow.

everyone else assumes that defining "_XOPEN_SOURCE" means XPG4.2,
so remove comparisions against the numerical values "3" or "4"

add RANDOM_MAX define from anon ymous

branches: 1.81.2;
Add csetexpandtc()

change some __attribute__s to __keywords

humanize_number was the only function in <util.h> that is not in libutil.
Move it to stdlib.h and add dehumanize_number.

Enable posix_memalign().

branches: 1.77.4;
The alloca() arg type check is broken (cpp do not do string comparisions)
so make it marginally less broken so that it works with other compilers
than gcc.
Probably the check can be removed, I doubt anyone will try to use gcc1
anymore.

Add strtof(3) and strtold(3); welcome libc 12.137.

u_intN_t -> uintN_t

__const__ -> const
This is in an __attribute__ so it isn't really necessary, but it gets
it out of the listing of __keyword__s...

Move getenv_r() to the local section.

Add getenv_r and ttyname_r

compat code reorg

Add a macro __MKTEMP_OK__ that renames mktemp() to _mktemp() for uses that
have been checked, so that the linker does not warn us. There are valid
uses for mktemp() such as for creating filenames that are going to be
targets of the rename system call. Discussed with lukem.

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

Add a function called getdevmajor().

It tells you the major device number for whatever character or block
device you ask it. This is sort of the inverse of devname(3) but not
quite, since it's backed by the kernel (sysctl's kern.drivers
information) and not a database cobbled together from the contents of
the filesystem.

Add unix/98 pty function prototypes.

more const sprinkling on cget* suggested by kleink.

constify the db name vector argument.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22270, verified by myself.

Do not give names to the parameters in function prototypes (those names
reduces the usefulness of the gcc -Wshadow flag).

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.

Well, _The_ Open Group.

Use the real TOG specification name for references.

1003.1-2001 has introduced unsetenv() which differs from the current
implementation in not permitting a "name=value" argument.
* Add a conforming __unsetenv13() and do function renaming for
unsetenv(); preserve old symbol with old behavior.
* Make visible setenv() and unsetenv() for 1003.1-2001 feature selection
macros; resolves PR standards/20479.

Add _Exit(), for C99 and POSIX-2001.
Proposed on tech-userlevel to a resounding silence in April last year.

add strsuftoll(3) and strsuftollx(3) - parse a string for a number
with an optional suffix, product (two numbers separated by `x'), and
provide range checking

add arc4random(3). from openbsd

branches: 1.55.2;
Don't declare qdiv() when qdiv_t wasn't defined previously; fixes PR lib/11557
from Richard Earnshaw.

Sprinkle some restrict qualifiers.

add prototypes for shquote() and shquotev()

branches: 1.52.2;
add prototypes for getprogname and setprogname

deal with bsearch duplicate declaration.

Move fmtcheck() prototype to stdio.h. It makes more sense to declare it
with the formatting functions with which it is meant to be used. Thanks
to Klaus Klein for "encouraging" me to make this change.

Add {__,}fmtcheck(), a function specified by Bill Sommerfeld to check
a user-supplied format string against a fallback format string to ensure
that they will consume arguments of the same type.

libc minor bump for new API entry point.

Allow format-arg checking for setproctitle

ISO/IEC 9899:1999: MB_CUR_MAX expands to a positive integer expression with
type size_t (not int).

(Incompatible ABI change, ignored due to a window of just two days.)

Preparation for the future introduction of multibyte locale.
- MB_LEN_MAX is increased to 32.
- To ensure binary compatibility for old executables
under multibyte locale, versioned setlocale is added.
- __mb_len_cur definision is added in setlocale.c
and enable it in stdlib.h .
It is also important for multibyte locale stuffs,
but I just forgot.

branches: 1.45.2; 1.45.4;
Add ISO C99 long long integer general utility interfaces; partially addresses
PR standards/9482.

Move <null.h> to <sys/null.h> for the purpose of avoiding conflicts due to
std C headers including it, and a kernel module written in C++ would benfit
as well.

For the benefit of G++'s null pointer constant implementation, #define NULL
as __null with egcs 1.0 (GCC 2.90) and above. As several headers are affected
by this change, move the definition into a new header file, <null.h>, to ease
maintenance.