Added comment with order of checks.

Supports 'yes', 'no' (default), and 'auto'.

Fix LDAP_API_VENDOR_VERSION -> LDAP_VENDOR_VERSION bug

AC_CYGWIN AC_MINGW32 AC_EXEEXT AC_OBJEXT

	Likely broke things for non-posix threadings....

Update -lldap_r implementation to:
	remove attribute support
	hide thread detachment
	provide concurrency accessors
	provide initialization function
	fix gethostby{addr,name}_r codes (not coverred by HAVE_REENTRANT_FUNCTIONS)
Update servers/libraries to use ldap_pvt_thread_ calls.
Cleanup server codes (no #ifdef HAVE_PTHREAD_THIS or _THATs)!
Removed -llthread

Rewrite all child waits to use wait4 (for consistency only).

libldap/string.c will hold various string.h replacements.
Removed liblutil/strdup.c
Moved ldap_pvt_strtok to string.c.
Moved prototype of ldap_pvt_strtok to <ac/string.h>.
Added #define strtok_r ldap_pvt_strtok (if strtok_r doesn't exist).

Suggest --with-threads=posix when error is reported.

Once we have private types for threads, this check can be removed.

strpbrk was checked for twice, and something didn't work. Now it does.

Improved ldap_int_strtok. If strtok_r does not exists, it will be worked
simulated with strspn and strpbrk. If strspn or strpbrk aren't there, they
will be supplied.
In addition, the memory-handling of the ldap_int_gethostby's is better now.

John's still working on the 'perl-test' (the perl backend test module).

in -lbind then in -lresolv.

creating them detached.  This hopefully will reduce problems on
draft4 pthread implementations related to creating detached
threads (which is _np under draft4) on some platforms without
causing problems with other thread implementations.

of a busy wait and allow the pthread_yield within the loop to be
removed.  This was the only pthread_yield which was necessary
when usng non-preemptive threads.  As such, the configure.in
sched_yield/pthread_yield missing error can be removed from
configure.in.  If explicit yield function is missing, just provide
a no-op replacement.
Moved declaration of various slapd globals from main.c to init.c
so they can be shared with ldbm tools.

  make depend, make tests, and make install all work when build directory
  is not the $srcdir.
Also modified library handling such that -lpthread more likely to be last.
WARNING: new orderring requires use of LDFLAGS to set global loader options
  such as -L/usr/local/lib.  If you put this in LIBS, some libraries
  may not be found a link time.
Likely broke Kerberos/LDAPD support.  Don't have those in my testbed.

one preprocessor macro instead of two.  Ran aclocal, autoconf,
autoheader to update derived files.  Modify util-int.c to use
CTIME_R_NARGS define and fallback to ctime() if not set.
Made LDAP_API_FEATURE_X_OPENLDAP_REENTRANT dependent upon result.

Teached autoconf to figure out how may arguments ctime_r expects.
Updated util-int.c to use this information.

Add check for to configure.in vsprintf, since HAVE_VSPRINTF is presumably
what was intended there.  Finally, also protected a bunch of if test
expressions with double quotes, as they were producing illegal test
commands.

Updated other codes as needed.

LDAP_SYSLOG in slap.h/slurp.h.

type 'sig_atomic_t' and marked it volatile.  Add detection to
configure to ensure sig_atomic_t and volatile are supported.

  -lldap_r can now be implemented...

including use of ctime_r, gethostby*_r, etc..  Also reworked ldap_sort_entries
to be reentrant.
Need to add code to properly set LDAP_API_FEATURE_THREAD_SAFE.

hosts_access depends on caller declaring some externs and, hence,
can be used for the test.

	Add plain -lpthread check.
	Move *_r() checks so that they are checked for regradless
		of thread libs used.  Before they were only being checked
		when using threads.
	Fix yields test such that compile/link/odd runtime failures results in
		'yes' instead of 'no'.

linking with $LTHREAD_LIBS.

Don't use new HAVE_FUNC_R on the client side (apps or libs).