[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]

[PATCH] 2038 fix



Attached is a patch to fix the 2038 bug in Perl.  It's pretty good.  There's
one obvious nit that needs more-expert-than-me attention.

* As noted by the #include "localtime64.c" in pp_sys.c, some sort of build foo
  has to be done to make things aware of that file.  I don't know how to do
  that.

I also eliminated the use of *S_my_localtime for the LOCALTIME_EDGECASE_BROKEN
special case because, well, it got in the way.  What's the deal with that?

I'm sure there's a few portability nits.  There's also a large pile of
optimizations planned (gmtime 2**56 takes a full minute).  Currently it runs
about 50% slower on OS X for small times.


PS  The underlying 64 bit time.h code is being developed independently here:
http://code.google.com/p/y2038/
there's lots of room for improvement and I'm a crappy C programmer.


-- 
94. Crucifixes do not ward off officers, and I should not test that.
    -- The 213 Things Skippy Is No Longer Allowed To Do In The U.S. Army
           http://skippyslist.com/list/

--- bleadperl/pp_sys.c	2008-06-28 14:11:23.000000000 -0700
+++ bleadperl.mv/pp_sys.c	2008-09-10 23:36:11.000000000 -0700
@@ -27,6 +27,8 @@
 #include "EXTERN.h"
 #define PERL_IN_PP_SYS_C
 #include "perl.h"
+#include "localtime64.h"
+#include "localtime64.c"
 
 #ifdef I_SHADOW
 /* Shadow password support for solaris - pdo@cs.umd.edu
@@ -4446,60 +4448,60 @@
 {
     dVAR;
     dSP;
-    Time_t when;
-    const struct tm *tmbuf;
+    Time64_T when;
+    struct tm tmbuf;
+    struct tm *err;
     static const char * const dayname[] =
 	{"Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"};
     static const char * const monname[] =
 	{"Jan", "Feb", "Mar", "Apr", "May", "Jun",
 	 "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"};
 
-    if (MAXARG < 1)
-	(void)time(&when);
+    if (MAXARG < 1) {
+	time_t now;
+	(void)time(&now);
+	when = (Time64_T)now;
+    }
     else
-#ifdef BIG_TIME
-	when = (Time_t)SvNVx(POPs);
-#else
-	when = (Time_t)SvIVx(POPs);
-#endif
+	when = (Time64_T)SvNVx(POPs);
 
     if (PL_op->op_type == OP_LOCALTIME)
-#ifdef LOCALTIME_EDGECASE_BROKEN
-	tmbuf = S_my_localtime(aTHX_ &when);
-#else
-	tmbuf = localtime(&when);
-#endif
+        err = localtime64_r(&when, &tmbuf);
     else
-	tmbuf = gmtime(&when);
+	err = gmtime64_r(&when, &tmbuf);
+
+    if( (tmbuf.tm_year + 1900) < 0 )
+	Perl_warner(aTHX_ packWARN(WARN_OVERFLOW),
+		    "local/gmtime under/overflowed the year");
 
     if (GIMME != G_ARRAY) {
 	SV *tsv;
         EXTEND(SP, 1);
         EXTEND_MORTAL(1);
-	if (!tmbuf)
+	if (err == NULL)
 	    RETPUSHUNDEF;
 	tsv = Perl_newSVpvf(aTHX_ "%s %s %2d %02d:%02d:%02d %d",
-			    dayname[tmbuf->tm_wday],
-			    monname[tmbuf->tm_mon],
-			    tmbuf->tm_mday,
-			    tmbuf->tm_hour,
-			    tmbuf->tm_min,
-			    tmbuf->tm_sec,
-			    tmbuf->tm_year + 1900);
+			    dayname[tmbuf.tm_wday],
+			    monname[tmbuf.tm_mon],
+			    tmbuf.tm_mday,
+			    tmbuf.tm_hour,
+			    tmbuf.tm_min,
+			    tmbuf.tm_sec,
+			    tmbuf.tm_year + 1900);
 	mPUSHs(tsv);
     }
-    else if (tmbuf) {
+    else if (err != NULL) {
         EXTEND(SP, 9);
         EXTEND_MORTAL(9);
-        mPUSHi(tmbuf->tm_sec);
-	mPUSHi(tmbuf->tm_min);
-	mPUSHi(tmbuf->tm_hour);
-	mPUSHi(tmbuf->tm_mday);
-	mPUSHi(tmbuf->tm_mon);
-	mPUSHi(tmbuf->tm_year);
-	mPUSHi(tmbuf->tm_wday);
-	mPUSHi(tmbuf->tm_yday);
-	mPUSHi(tmbuf->tm_isdst);
+        mPUSHi(tmbuf.tm_sec);
+	mPUSHi(tmbuf.tm_min);
+	mPUSHi(tmbuf.tm_hour);
+	mPUSHi(tmbuf.tm_mday);
+	mPUSHi(tmbuf.tm_mon);
+	mPUSHi(tmbuf.tm_year);
+	mPUSHi(tmbuf.tm_wday);
+	mPUSHi(tmbuf.tm_yday);
+	mPUSHi(tmbuf.tm_isdst);
     }
     RETURN;
 }
--- /dev/null	2008-09-11 00:01:41.000000000 -0700
+++ bleadperl.mv/localtime64.c	2008-09-10 23:35:51.000000000 -0700
@@ -0,0 +1,238 @@
+/* localtime64.c */
+
+#include <assert.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include <time.h>
+
+static const int days_in_month[2][12] = {
+    {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31},
+    {31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31},
+};
+
+static const int julian_days_by_month[2][12] = {
+    {0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334},
+    {0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335},
+};
+
+static const int length_of_year[2] = { 365, 366 };
+
+/* 28 year calendar cycle between 2010 and 2037 */
+static const int safe_years[28] = {
+    2016, 2017, 2018, 2019,
+    2020, 2021, 2022, 2023,
+    2024, 2025, 2026, 2027,
+    2028, 2029, 2030, 2031,
+    2032, 2033, 2034, 2035,
+    2036, 2037, 2010, 2011,
+    2012, 2013, 2014, 2015
+};
+
+static const int dow_year_start[28] = {
+    5, 0, 1, 2,     /* 2016 - 2019 */
+    3, 5, 6, 0,
+    1, 3, 4, 5,
+    6, 1, 2, 3,
+    4, 6, 0, 1,
+    2, 4, 5, 6,     /* 2036, 2037, 2010, 2011 */
+    0, 2, 3, 4      /* 2012, 2013, 2014, 2015 */
+};
+
+
+#define IS_LEAP(n)	((!(((n) + 1900) % 400) || (!(((n) + 1900) % 4) && (((n) + 1900) % 100))) != 0)
+#define WRAP(a,b,m)	((a) = ((a) <  0  ) ? ((b)--, (a) + (m)) : (a))
+
+int _is_exception_century(long year)
+{
+    int is_exception = ((year % 100 == 0) && !(year % 400 == 0));
+    /* printf("is_exception_century: %s\n", is_exception ? "yes" : "no"); */
+
+    return(is_exception);
+}
+
+void _check_tm(struct tm *tm)
+{
+    /* Don't forget leap seconds */
+    assert(tm->tm_sec  >= 0 && tm->tm_sec <= 61);
+    assert(tm->tm_min  >= 0 && tm->tm_min <= 59);
+    assert(tm->tm_hour >= 0 && tm->tm_hour <= 23);
+    assert(tm->tm_mday >= 1 && tm->tm_mday <= 31);
+    assert(tm->tm_mon  >= 0 && tm->tm_mon  <= 11);
+    assert(tm->tm_wday >= 0 && tm->tm_wday <= 6);
+    assert(tm->tm_yday >= 0 && tm->tm_yday <= 365);
+
+#ifdef TM_HAS_GMTOFF
+    assert(   tm->tm_gmtoff >= -24 * 60 * 60
+           && tm->tm_gmtoff <=  24 * 60 * 60);
+#endif
+
+    if( !IS_LEAP(tm->tm_year) ) {
+        /* no more than 365 days in a non_leap year */
+        assert( tm->tm_yday <= 364 );
+
+        /* and no more than 28 days in Feb */
+        if( tm->tm_mon == 1 ) {
+            assert( tm->tm_mday <= 28 );
+        }
+    }
+}
+
+/* The exceptional centuries without leap years cause the cycle to
+   shift by 16
+*/
+int _cycle_offset(long year)
+{
+    const long start_year = 2000;
+    long year_diff  = year - start_year - 1;
+    long exceptions  = year_diff / 100;
+    exceptions     -= year_diff / 400;
+
+    assert( year >= 2001 );
+
+    /* printf("year: %d, exceptions: %d\n", year, exceptions); */
+
+    return exceptions * 16;
+}
+
+/* For a given year after 2038, pick the latest possible matching
+   year in the 28 year calendar cycle.
+*/
+#define SOLAR_CYCLE_LENGTH 28
+int _safe_year(long year)
+{
+    int safe_year;
+    long year_cycle = year + _cycle_offset(year);
+
+    /* Change non-leap xx00 years to an equivalent */
+    if( _is_exception_century(year) )
+        year_cycle += 11;
+    
+    year_cycle %= SOLAR_CYCLE_LENGTH;
+    
+    safe_year = safe_years[year_cycle];
+
+    assert(safe_year <= 2037 && safe_year >= 2010);
+    
+    /*
+    printf("year: %d, year_cycle: %d, safe_year: %d\n",
+           year, year_cycle, safe_year);
+    */
+
+    return safe_year;
+}
+
+struct tm *gmtime64_r (const Time64_T *in_time, struct tm *p)
+{
+    int v_tm_sec, v_tm_min, v_tm_hour, v_tm_mon, v_tm_wday;
+    Time64_T v_tm_tday;
+    int leap;
+    Time64_T m;
+    Time64_T time = *in_time;
+    
+#ifdef TM_HAS_GMTOFF
+    p->tm_gmtoff = 0;
+#endif
+    p->tm_isdst  = 0;
+
+#ifdef TM_HAS_ZONE
+    p->tm_zone   = "UTC";
+#endif
+    
+    v_tm_sec =  time % 60;
+    time /= 60;
+    v_tm_min =  time % 60;
+    time /= 60;
+    v_tm_hour = time % 24;
+    time /= 24;
+    v_tm_tday = time;
+    WRAP (v_tm_sec, v_tm_min, 60);
+    WRAP (v_tm_min, v_tm_hour, 60);
+    WRAP (v_tm_hour, v_tm_tday, 24);
+    if ((v_tm_wday = (v_tm_tday + 4) % 7) < 0)
+        v_tm_wday += 7;
+    m = v_tm_tday;
+    if (m >= 0) {
+        p->tm_year = 70;
+        leap = IS_LEAP (p->tm_year);
+        while (m >= (Time64_T) length_of_year[leap]) {
+            m -= (Time64_T) length_of_year[leap];
+            p->tm_year++;
+            leap = IS_LEAP (p->tm_year);
+        }
+        v_tm_mon = 0;
+        while (m >= (Time64_T) days_in_month[leap][v_tm_mon]) {
+            m -= (Time64_T) days_in_month[leap][v_tm_mon];
+            v_tm_mon++;
+        }
+    } else {
+        p->tm_year = 69;
+        leap = IS_LEAP (p->tm_year);
+        while (m < (Time64_T) -length_of_year[leap]) {
+            m += (Time64_T) length_of_year[leap];
+            p->tm_year--;
+            leap = IS_LEAP (p->tm_year);
+        }
+        v_tm_mon = 11;
+        while (m < (Time64_T) -days_in_month[leap][v_tm_mon]) {
+            m += (Time64_T) days_in_month[leap][v_tm_mon];
+            v_tm_mon--;
+        }
+        m += (Time64_T) days_in_month[leap][v_tm_mon];
+    }
+    p->tm_mday = (int) m + 1;
+    p->tm_yday = julian_days_by_month[leap][v_tm_mon] + m;
+    p->tm_sec = v_tm_sec, p->tm_min = v_tm_min, p->tm_hour = v_tm_hour,
+        p->tm_mon = v_tm_mon, p->tm_wday = v_tm_wday;
+    
+    _check_tm(p);
+
+    return p;
+}
+
+
+struct tm *localtime64_r (const Time64_T *time, struct tm *local_tm)
+{
+    time_t safe_time;
+    struct tm gm_tm;
+    long orig_year;
+    int month_diff;
+
+    gmtime64_r(time, &gm_tm);
+    orig_year = gm_tm.tm_year;
+
+    if (gm_tm.tm_year > (2037 - 1900))
+        gm_tm.tm_year = _safe_year(gm_tm.tm_year + 1900) - 1900;
+
+    safe_time = timegm(&gm_tm);
+    localtime_r(&safe_time, local_tm);
+
+    local_tm->tm_year = orig_year;
+    month_diff = local_tm->tm_mon - gm_tm.tm_mon;
+
+    /*  When localtime is Dec 31st previous year and
+        gmtime is Jan 1st next year.
+    */
+    if( month_diff == 11 ) {
+        local_tm->tm_year--;
+    }
+
+    /*  When localtime is Jan 1st, next year and
+        gmtime is Dec 31st, previous year.
+    */
+    if( month_diff == -11 ) {
+        local_tm->tm_year++;
+    }
+
+    /* GMT is Jan 1st, xx01 year, but localtime is still Dec 31st 
+       in a non-leap xx00.  There is one point in the cycle
+       we can't account for which the safe xx00 year is a leap
+       year.  So we need to correct for Dec 31st comming out as
+       the 366th day of the year.
+    */
+    if( !IS_LEAP(local_tm->tm_year) && local_tm->tm_yday == 365 )
+        local_tm->tm_yday--;
+
+    _check_tm(local_tm);
+    
+    return local_tm;
+}
--- /dev/null	2008-09-11 00:01:41.000000000 -0700
+++ bleadperl.mv/localtime64.h	2008-09-10 21:56:11.000000000 -0700
@@ -0,0 +1,11 @@
+#include <time.h>
+
+#ifndef LOCALTIME64_H
+#    define LOCALTIME64_H
+
+typedef Quad_t Time64_T;
+
+struct tm *gmtime64_r    (const Time64_T *, struct tm *);
+struct tm *localtime64_r (const Time64_T *, struct tm *);
+
+#endif
--- bleadperl/t/op/time.t	2007-01-07 16:18:22.000000000 -0800
+++ bleadperl.mv/t/op/time.t	2008-09-10 23:57:44.000000000 -0700
@@ -1,14 +1,12 @@
 #!./perl
 
-$does_gmtime = gmtime(time);
-
 BEGIN {
     chdir 't' if -d 't';
     @INC = '../lib';
     require './test.pl';
 }
 
-plan tests => 8;
+plan tests => 32;
 
 ($beguser,$begsys) = times;
 
@@ -32,7 +30,9 @@
 ($xsec,$foo) = localtime($now);
 $localyday = $yday;
 
-ok($sec != $xsec && $mday && $year,             'localtime() list context');
+isnt($sec, $xsec),      'localtime() list context';
+ok $mday,               '  month day';
+ok $year,               '  year';
 
 ok(localtime() =~ /^(Sun|Mon|Tue|Wed|Thu|Fri|Sat)[ ]
                     (Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)[ ]
@@ -56,13 +56,13 @@
 ok($hour != $hour2,                             'changes to $ENV{TZ} respected');
 }
 
-SKIP: {
-    skip "No gmtime()", 3 unless $does_gmtime;
 
 ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,$isdst) = gmtime($beg);
 ($xsec,$foo) = localtime($now);
 
-ok($sec != $xsec && $mday && $year,             'gmtime() list context');
+isnt($sec, $xsec),      'gmtime() list conext';
+ok $mday,               '  month day';
+ok $year,               '  year';
 
 my $day_diff = $localyday - $yday;
 ok( grep({ $day_diff == $_ } (0, 1, -1, 364, 365, -364, -365)),
@@ -76,4 +76,49 @@
                /x,
    'gmtime(), scalar context'
   );
+
+
+
+# Test gmtime over a range of times.
+{
+    # gm/localtime is limited by the size of tm_year which might be as small as 16 bits
+    my %tests = (
+        # time_t         gmtime list                          scalar
+        -2**35        => [52, 13, 20, 7, 2, -1019, 5, 65, 0, "Fri Mar  7 20:13:52 881"],
+        -2**32        => [44, 31, 17, 24, 10, -67, 0, 327, 0, "Sun Nov 24 17:31:44 1833"],
+        -2**31        => [52, 45, 20, 13, 11, 1, 5, 346, 0, "Fri Dec 13 20:45:52 1901"],
+        0             => [0, 0, 0, 1, 0, 70, 4, 0, 0, "Thu Jan  1 00:00:00 1970"],
+        2**30         => [4, 37, 13, 10, 0, 104, 6, 9, 0, "Sat Jan 10 13:37:04 2004"],
+        2**31         => [8, 14, 3, 19, 0, 138, 2, 18, 0, "Tue Jan 19 03:14:08 2038"],
+        2**32         => [16, 28, 6, 7, 1, 206, 0, 37, 0, "Sun Feb  7 06:28:16 2106"],
+        2**39         => [8, 18, 12, 25, 0, 17491, 2, 24, 0, "Tue Jan 25 12:18:08 19391"],
+    );
+
+    for my $time (keys %tests) {
+        my @expected  = @{$tests{$time}};
+        my $scalar    = pop @expected;
+
+        ok eq_array([gmtime($time)], \@expected),  "gmtime($time) list context";
+        is scalar gmtime($time), $scalar,       "  scalar";
+    }
+}
+
+
+# Test localtime
+{
+    # We pick times which fall in the middle of a month, so the month and year should be
+    # the same regardless of the time zone.
+    my %tests = (
+        # time_t           month, year,  scalar
+        5000000000      => [5,  228,     qr/Jun \d+ .* 2128$/],
+        1163500000      => [10, 106,     qr/Nov \d+ .* 2006$/],
+    );
+
+    for my $time (keys %tests) {
+        my @expected  = @{$tests{$time}};
+        my $scalar    = pop @expected;
+
+        ok eq_array([(localtime($time))[4,5]], \@expected),  "localtime($time) list context";
+        like scalar localtime($time), $scalar,       "  scalar";
+    }
 }

Follow-Ups from:
Michael G Schwern <schwern@pobox.com>

[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]