- tz - lists.iana.org

New timezone Antarctica/Concordia (UTC+8)
by Erik Geletti Nov. 26, 2024

Nov. 26, 2024

Goodmorning. I'm writing here to propose a new time zone for an all-year open Antarctic base. The base is a French–Italian research facility that was built 3,233 m (10,607 ft) above sea level at a location called Dome C on the Antarctic Plateau, Antarctica. https://en.wikipedia.org/wiki/Concordia_Station The timezone is UTC+8 without DST. Best regards, Erik Geletti -- Erik Geletti Servizi Tecnici ed Informatici Istituto Nazionale di Oceanografia e di Geofisica Sperimentale - OGS Borgo Grotta Gigante n. 42/c 34010 Sgonico (Trieste) - Italia Tel. 0402140286 www.ogs.it

8 17

[PROPOSED 1/2] Move iinntt definition
by Paul Eggert Nov. 24, 2024

Nov. 24, 2024

It’s used only in localtime.c, so move its definition there. * localtime.c (IINNTT_MIN, IINNTT_MAX, iinntt): Move to here ... * private.h: ... from here. --- localtime.c | 18 ++++++++++++++++++ private.h | 17 ----------------- 2 files changed, 18 insertions(+), 17 deletions(-) diff --git a/localtime.c b/localtime.c index 14c356e1..1d42fc36 100644 --- a/localtime.c +++ b/localtime.c @@ -29,6 +29,24 @@ static int lock(void) { return 0; } static void unlock(void) { } #endif +/* A signed type wider than int, so that we can add 1900 + tm_mon/12 to tm_year + without overflow. The static_assert checks that it is indeed wider + than int; if this fails on your platform please let us know. */ +#if INT_MAX < LONG_MAX +typedef long iinntt; +# define IINNTT_MIN LONG_MIN +# define IINNTT_MAX LONG_MAX +#elif INT_MAX < LLONG_MAX +typedef long long iinntt; +# define IINNTT_MIN LLONG_MIN +# define IINNTT_MAX LLONG_MAX +#else +typedef intmax_t iinntt; +# define IINNTT_MIN INTMAX_MIN +# define IINNTT_MAX INTMAX_MAX +#endif +static_assert(IINNTT_MIN < INT_MIN && INT_MAX < IINNTT_MAX); + /* On platforms where offtime or mktime might overflow, strftime.c defines USE_TIMEX_T to be true and includes us. This tells us to #define time_t to an internal type timex_t that is diff --git a/private.h b/private.h index 3a699f0f..bdd00013 100644 --- a/private.h +++ b/private.h @@ -450,23 +450,6 @@ typedef unsigned long uintmax_t; #endif /* PORT_TO_C89 */ -/* A signed type wider than int, so that we can add 1900 to tm_year - without overflow. */ -#if INT_MAX < LONG_MAX -typedef long iinntt; -# define IINNTT_MIN LONG_MIN -# define IINNTT_MAX LONG_MAX -#elif INT_MAX < LLONG_MAX -typedef long long iinntt; -# define IINNTT_MIN LLONG_MIN -# define IINNTT_MAX LLONG_MAX -#else -typedef intmax_t iinntt; -# define IINNTT_MIN INTMAX_MIN -# define IINNTT_MAX INTMAX_MAX -#endif -static_assert(IINNTT_MIN < INT_MIN && INT_MAX < IINNTT_MAX); - /* The maximum size of any created object, as a signed integer. Although the C standard does not outright prohibit larger objects, behavior is undefined if the result of pointer subtraction does not -- 2.43.0

1 1

Re: [PROPOSED 4/6] Fix bug near the year 2**31 - 1 - 1900
by Paul Eggert Nov. 23, 2024

Nov. 23, 2024

On 2024-11-22 22:08, Magnus Fromreide wrote: > The definition of iinntt does not guarantee that it is wider than an int, > does something break if INT_MAX == INTMAX_MAX and INT_MIN == INTMAX_MIN? That's checked statically by "static_assert(IINNTT_MIN < INT_MIN && INT_MAX < IINNTT_MAX);". I don't know of any tzcode platform where that static assertion would fail (and thus break the build) because no signed type is wider than 'int'. If you know of one, please let us know.

1 0

(unofficial) Australian Central Western Time
by Stuart Bishop Nov. 22, 2024

Nov. 22, 2024

An article discussing the in-use but technically unofficial timezones in the Western Australian portion of the Nullarbor Plain. https://www.abc.net.au/news/2024-11-22/outback-wa-properties-strange-time-z… -- Stuart Bishop <stuart(a)stuartbishop.net> http://www.stuartbishop.net/

2 1

[PROPOSED] Avoid time_t + int overflow
by Paul Eggert Nov. 22, 2024

Nov. 22, 2024

* localtime.c (increment_overflow_time_int): New function. (time2sub): Use it (instead of relying on possibly-undefined behavior) to check for time_t + int overflow. --- localtime.c | 19 ++++++++++++++++--- 1 file changed, 16 insertions(+), 3 deletions(-) diff --git a/localtime.c b/localtime.c index 07ce41b8..14c356e1 100644 --- a/localtime.c +++ b/localtime.c @@ -1882,6 +1882,21 @@ increment_overflow_iinntt(iinntt *lp, int m) #endif } +static bool +increment_overflow_time_int(time_t *tp, int j) +{ +#ifdef ckd_add + return ckd_add(tp, *tp, j); +#else + if (j < 0 + ? (TYPE_SIGNED(time_t) ? *tp < TIME_T_MIN - j : *tp <= -1 - j) + : TIME_T_MAX - j < *tp) + return true; + *tp += j; + return false; +#endif +} + static bool increment_overflow_time(time_t *tp, int_fast32_t j) { @@ -2155,10 +2170,8 @@ time2sub(struct tm *const tmp, return WRONG; } label: - newt = t + saved_seconds; - if ((newt < t) != (saved_seconds < 0)) + if (increment_overflow_time_int(&t, saved_seconds)) return WRONG; - t = newt; if (funcp(sp, &t, offset, tmp)) *okayp = true; return t; -- 2.47.0

1 0

Dead code in strftime()?
by Tom Lane Nov. 22, 2024

Nov. 22, 2024

strftime() in the 2024b tzcode distribution includes if (!p) { errno = EOVERFLOW; return 0; } AFAICS this is dead code, except maybe in the case that the call passes s = NULL, maxsize = 0, which doesn't seem like a case that the code intends to support (and EOVERFLOW would be a rather odd choice of error code if that's the idea). newstrftime.3.txt offers a clue about the intent: This function may fail if: [EOVERFLOW] The format includes an %s conversion and the number of seconds since the Epoch cannot be represented in a time_t. But the %s conversion code never returns zero. There's a suggestive comment there: /* If mktime fails, %s expands to the value of (time_t) -1 as a failure marker; this is better in practice than strftime failing. */ So it looks to me like somebody changed the error handling in the %s stanza but didn't bother to remove the now-dead path in the caller nor update the man page. Am I missing something? regards, tom lane

3 13

[PROPOSED 1/6] Improve zdump overflow checking
by Paul Eggert Nov. 22, 2024

Nov. 22, 2024

This also pacifies -Wsuggest-attribute=pure on x86 on platforms that lack tm_gmtoff and tm_zone. * zdump.c (xstrsize): Remove; no longer used. (tzalloc, saveabbr): Use plain strlen here, since the string sizes have already been checked. (main): Limit ‘longest’ to INT_MAX - 2 so that a printf with it and two spaces does not flake out. Check for arglen overflow once here, rather than on every alloc. --- zdump.c | 24 ++++++++---------------- 1 file changed, 8 insertions(+), 16 deletions(-) diff --git a/zdump.c b/zdump.c index e8178733..b668022d 100644 --- a/zdump.c +++ b/zdump.c @@ -148,17 +148,6 @@ sumsize(ptrdiff_t a, ptrdiff_t b) size_overflow(); } -/* Return the size of of the string STR, including its trailing NUL. - Report an error and exit if this would exceed INDEX_MAX which means - pointer subtraction wouldn't work. */ -static ptrdiff_t -xstrsize(char const *str) -{ - size_t len = strlen(str); - if (len < INDEX_MAX) - return len + 1; - size_overflow(); -} /* Return a pointer to a newly allocated buffer of size SIZE, exiting on failure. SIZE should be positive. */ @@ -266,7 +255,7 @@ tzalloc(char const *val) static ptrdiff_t fakeenv0size; void *freeable = NULL; char **env = fakeenv, **initial_environ; - ptrdiff_t valsize = xstrsize(val); + ptrdiff_t valsize = strlen(val) + 1; if (fakeenv0size < valsize) { char **e = environ, **to; ptrdiff_t initial_nenvptrs = 1; /* Counting the trailing NULL pointer. */ @@ -425,7 +414,7 @@ saveabbr(char **buf, ptrdiff_t *bufalloc, struct tm const *tmp) if (HAVE_LOCALTIME_RZ) return ab; else { - ptrdiff_t absize = xstrsize(ab); + ptrdiff_t absize = strlen(ab) + 1; if (*bufalloc < absize) { free(*buf); @@ -487,6 +476,7 @@ main(int argc, char *argv[]) register time_t cuthitime; time_t now; bool iflag = false; + size_t arglenmax = 0; cutlotime = absolute_min_time; cuthitime = absolute_max_time; @@ -586,12 +576,14 @@ main(int argc, char *argv[]) now = time(NULL); now |= !now; } - longest = 0; for (i = optind; i < argc; i++) { size_t arglen = strlen(argv[i]); - if (longest < arglen) - longest = min(arglen, INT_MAX); + if (arglenmax < arglen) + arglenmax = arglen; } + if (!HAVE_SETENV && INDEX_MAX <= arglenmax) + size_overflow(); + longest = min(arglenmax, INT_MAX - 2); for (i = optind; i < argc; ++i) { timezone_t tz = tzalloc(argv[i]); -- 2.47.0

1 5

(no subject)
by Cool Gab19 Nov. 8, 2024

Nov. 8, 2024

can you add Philippine Daylight Time on your Database I think it was missing from the list of the time zones please reply immediately if solved thanks...

4 8

[PROPOSED] * NEWS, europe: Belated thanks for Ittoqqortoormiit 2024 fix
by Paul Eggert Nov. 8, 2024

Nov. 8, 2024

--- NEWS | 2 +- europe | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/NEWS b/NEWS index a109ae9d..8381a197 100644 --- a/NEWS +++ b/NEWS @@ -233,7 +233,7 @@ Release 2023d - 2023-12-21 20:02:24 -0800 changing its time zone from -01/+00 to -02/-01 at the same moment as the spring-forward transition. Its clocks will therefore not spring forward as previously scheduled. The time zone change - reverts to its common practice before 1981. + reverts to its common practice before 1981. (Thanks to Jule Dabars.) Fix predictions for DST transitions in Palestine in 2072-2075, correcting a typo introduced in 2023a. (Thanks to Jule Dabars.) diff --git a/europe b/europe index f9063949..df334fc2 100644 --- a/europe +++ b/europe @@ -1147,7 +1147,7 @@ Zone Atlantic/Faroe -0:27:04 - LMT 1908 Jan 11 # Tórshavn # However, Greenland will change to Daylight Saving Time again in 2024 # and onwards. -# From a contributor who wishes to remain anonymous for now (2023-10-29): +# From Jule Dabars (2023-10-29): # https://www.dr.dk/nyheder/seneste/i-nat-skal-uret-stilles-en-time-tilbage-m… # with a link to that page: # https://naalakkersuisut.gl/Nyheder/2023/10/2710_sommertid -- 2.43.0

1 0

[PROPOSED] * NEWS: Belated thanks for Palestine 2072-2075 fix
by Paul Eggert Nov. 7, 2024

Nov. 7, 2024

--- NEWS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/NEWS b/NEWS index 8c315643..a109ae9d 100644 --- a/NEWS +++ b/NEWS @@ -236,7 +236,7 @@ Release 2023d - 2023-12-21 20:02:24 -0800 reverts to its common practice before 1981. Fix predictions for DST transitions in Palestine in 2072-2075, - correcting a typo introduced in 2023a. + correcting a typo introduced in 2023a. (Thanks to Jule Dabars.) Changes to past and future timestamps -- 2.43.0

1 0