MFV of tzcode2009e:

Upgrade of the tzcode from 2004a to 2009e.

Changes are numerous, but include...

- New format of the output of zic, which supports both 32 and 64
  bit time_t formats.

- zdump on 64 bit platforms will actually produce some output instead
  of doing nothing for a looooooooong time.

- linux_base-fX, with X >= at least 8, will work without problems related
  to the local time again.

The original patch, based on the 2008e, has been running for a long
time on both my laptop and desktop machine and have been tested by
other people.

After the installation of this code and the running of zic(8), you
need to run tzsetup(8) again to install the new datafile.

Approved by:	wollman@ for usr.sbin/zic
MFC after:	1 month

1 files changed, 40 insertions, 58 deletions

diff --git a/lib/libc/stdtime/difftime.c b/lib/libc/stdtime/difftime.c
index 596ea82..d16f9a0 100644
--- a/lib/libc/stdtime/difftime.c
+++ b/lib/libc/stdtime/difftime.c
@@ -1,12 +1,12 @@
 /*
 ** This file is in the public domain, so clarified as of
-** June 5, 1996 by Arthur David Olson (arthur_david_olson@nih.gov).
+** 1996-06-05 by Arthur David Olson.
 */
 
 #include <sys/cdefs.h>
 #ifndef lint
 #ifndef NOID
-static char	elsieid[] __unused = "@(#)difftime.c	7.9";
+static char	elsieid[] __unused = "@(#)difftime.c	8.1";
 #endif /* !defined NOID */
 #endif /* !defined lint */
 __FBSDID("$FreeBSD$");
@@ -14,74 +14,56 @@ __FBSDID("$FreeBSD$");
 /*LINTLIBRARY*/
 
 #include "namespace.h"
-#include "private.h"
+#include "private.h"	/* for time_t, TYPE_INTEGRAL, and TYPE_SIGNED */
 #include "un-namespace.h"
 
-/*
-** Algorithm courtesy Paul Eggert (eggert@twinsun.com).
-*/
-
-#ifdef HAVE_LONG_DOUBLE
-#define long_double	long double
-#endif /* defined HAVE_LONG_DOUBLE */
-#ifndef HAVE_LONG_DOUBLE
-#define long_double	double
-#endif /* !defined HAVE_LONG_DOUBLE */
-
 double
 difftime(time1, time0)
 const time_t	time1;
 const time_t	time0;
 {
-	time_t	delta;
-	time_t	hibit;
-
-	{
-		time_t		tt;
-		double		d;
-		long_double	ld;
-
-		if (sizeof tt < sizeof d)
-			return (double) time1 - (double) time0;
-		if (sizeof tt < sizeof ld)
-			return (long_double) time1 - (long_double) time0;
+	/*
+	** If (sizeof (double) > sizeof (time_t)) simply convert and subtract
+	** (assuming that the larger type has more precision).
+	** This is the common real-world case circa 2004.
+	*/
+	if (sizeof (double) > sizeof (time_t))
+		return (double) time1 - (double) time0;
+	if (!TYPE_INTEGRAL(time_t)) {
+		/*
+		** time_t is floating.
+		*/
+		return time1 - time0;
+	}
+	if (!TYPE_SIGNED(time_t)) {
+		/*
+		** time_t is integral and unsigned.
+		** The difference of two unsigned values can't overflow
+		** if the minuend is greater than or equal to the subtrahend.
+		*/
+		if (time1 >= time0)
+			return time1 - time0;
+		else	return -((double) (time0 - time1));
 	}
-	if (time1 < time0)
-		return -difftime(time0, time1);
 	/*
-	** As much as possible, avoid loss of precision
-	** by computing the difference before converting to double.
+	** time_t is integral and signed.
+	** Handle cases where both time1 and time0 have the same sign
+	** (meaning that their difference cannot overflow).
 	*/
-	delta = time1 - time0;
-	if (delta >= 0)
-		return delta;
+	if ((time1 < 0) == (time0 < 0))
+		return time1 - time0;
 	/*
-	** Repair delta overflow.
+	** time1 and time0 have opposite signs.
+	** Punt if unsigned long is too narrow.
 	*/
-	hibit = (~ (time_t) 0) << (TYPE_BIT(time_t) - 1);
+	if (sizeof (unsigned long) < sizeof (time_t))
+		return (double) time1 - (double) time0;
 	/*
-	** The following expression rounds twice, which means
-	** the result may not be the closest to the true answer.
-	** For example, suppose time_t is 64-bit signed int,
-	** long_double is IEEE 754 double with default rounding,
-	** time1 = 9223372036854775807 and time0 = -1536.
-	** Then the true difference is 9223372036854777343,
-	** which rounds to 9223372036854777856
-	** with a total error of 513.
-	** But delta overflows to -9223372036854774273,
-	** which rounds to -9223372036854774784, and correcting
-	** this by subtracting 2 * (long_double) hibit
-	** (i.e. by adding 2**64 = 18446744073709551616)
-	** yields 9223372036854776832, which
-	** rounds to 9223372036854775808
-	** with a total error of 1535 instead.
-	** This problem occurs only with very large differences.
-	** It's too painful to fix this portably.
-	** We are not alone in this problem;
-	** some C compilers round twice when converting
-	** large unsigned types to small floating types,
-	** so if time_t is unsigned the "return delta" above
-	** has the same double-rounding problem with those compilers.
+	** Stay calm...decent optimizers will eliminate the complexity below.
 	*/
-	return delta - 2 * (long_double) hibit;
+	if (time1 >= 0 /* && time0 < 0 */)
+		return (unsigned long) time1 +
+			(unsigned long) (-(time0 + 1)) + 1;
+	return -(double) ((unsigned long) time0 +
+		(unsigned long) (-(time1 + 1)) + 1);
 }