- tz - lists.iana.org

FW: FW: stack overflow in tzload
by Olson, Arthur David (NIH/NCI) [E] Nov. 23, 2010

Nov. 23, 2010

I'm forwarding this message from Ted Unangst, who is not on the time zone mailing list. Those of you who are on the list, please direct replies appropriately. --ado -----Original Message----- From: Ted Unangst [mailto:ted.unangst@gmail.com] Sent: Tuesday, November 23, 2010 10:08 To: Christos Zoulas Cc: tz(a)elsie.nci.nih.gov; tz(a)lecserver.nci.nih.gov; Olson, Arthur David (NIH/NCI) [E] Subject: Re: FW: stack overflow in tzload On Mon, Nov 22, 2010 at 11:28 PM, Christos Zoulas <christos(a)zoulas.com> wrote: > On Nov 22, 4:44pm, eggert(a)cs.ucla.edu (Paul Eggert) wrote: > -- Subject: Re: FW: stack overflow in tzload > > | Surely this sort of change should be put in only conditionally. > | On most machines, the change would will slow performance down due to the > | malloc overhead and would add one more point of failure. > | Only on machines with tiny stacks is the change a win. > > Sure, but these days the cost of malloc is really small, and having more > ifdefs in the code just cause more maintenance. So we either adopt the > change, or we just tell Ted that if you run with < 256K of stack you > can expect lots of things to break - not just that - so give yourself more All fair points, but it happens that this is the thing that broke. Smallish stacks are not uncommon if you expect to be creating many threads (the app in question is a of web server not of my own design, I'm not even sure what stack size it expects). The real wrinkle is that the tz code is part of the system library, I expect it to just work. If somebody had asked last week "How much stack does calling gmtime() require?", I suspect most people's answers would fall short of the mark, apart from maybe a very few people on this list. We can just patch this in openbsd, it's not that big a deal, but wanted to pass the change upstream as well.

1 0

stack overflow in tzload
by Ted Unangst Nov. 22, 2010

Nov. 22, 2010

Stack overflow as in there's not enough stack. I ran into this problem running a program on openbsd that was using a fairly small thread size. I switched to using malloc. Index: localtime.c =================================================================== RCS file: /home/tedu/cvs/src/lib/libc/time/localtime.c,v retrieving revision 1.35 diff -u -r1.35 localtime.c --- localtime.c 23 Aug 2010 22:35:34 -0000 1.35 +++ localtime.c 20 Nov 2010 23:32:09 -0000 @@ -340,7 +340,7 @@ char buf[2 * sizeof(struct tzhead) + 2 * sizeof *sp + 4 * TZ_MAX_TIMES]; - } u; + } *u; sp->goback = sp->goahead = FALSE; if (name != NULL && issetugid() != 0) @@ -383,28 +383,33 @@ if ((fid = open(name, OPEN_MODE)) == -1) return -1; } - nread = read(fid, u.buf, sizeof u.buf); - if (close(fid) < 0 || nread <= 0) + u = malloc(sizeof(*u)); + if (!u) { + close(fid); return -1; + } + nread = read(fid, u->buf, sizeof u->buf); + if (close(fid) < 0 || nread <= 0) + goto bad; for (stored = 4; stored <= 8; stored *= 2) { int ttisstdcnt; int ttisgmtcnt; - ttisstdcnt = (int) detzcode(u.tzhead.tzh_ttisstdcnt); - ttisgmtcnt = (int) detzcode(u.tzhead.tzh_ttisgmtcnt); - sp->leapcnt = (int) detzcode(u.tzhead.tzh_leapcnt); - sp->timecnt = (int) detzcode(u.tzhead.tzh_timecnt); - sp->typecnt = (int) detzcode(u.tzhead.tzh_typecnt); - sp->charcnt = (int) detzcode(u.tzhead.tzh_charcnt); - p = u.tzhead.tzh_charcnt + sizeof u.tzhead.tzh_charcnt; + ttisstdcnt = (int) detzcode(u->tzhead.tzh_ttisstdcnt); + ttisgmtcnt = (int) detzcode(u->tzhead.tzh_ttisgmtcnt); + sp->leapcnt = (int) detzcode(u->tzhead.tzh_leapcnt); + sp->timecnt = (int) detzcode(u->tzhead.tzh_timecnt); + sp->typecnt = (int) detzcode(u->tzhead.tzh_typecnt); + sp->charcnt = (int) detzcode(u->tzhead.tzh_charcnt); + p = u->tzhead.tzh_charcnt + sizeof u->tzhead.tzh_charcnt; if (sp->leapcnt < 0 || sp->leapcnt > TZ_MAX_LEAPS || sp->typecnt <= 0 || sp->typecnt > TZ_MAX_TYPES || sp->timecnt < 0 || sp->timecnt > TZ_MAX_TIMES || sp->charcnt < 0 || sp->charcnt > TZ_MAX_CHARS || (ttisstdcnt != sp->typecnt && ttisstdcnt != 0) || (ttisgmtcnt != sp->typecnt && ttisgmtcnt != 0)) - return -1; - if (nread - (p - u.buf) < + goto bad; + if (nread - (p - u->buf) < sp->timecnt * stored + /* ats */ sp->timecnt + /* types */ sp->typecnt * 6 + /* ttinfos */ @@ -412,7 +417,7 @@ sp->leapcnt * (stored + 4) + /* lsinfos */ ttisstdcnt + /* ttisstds */ ttisgmtcnt) /* ttisgmts */ - return -1; + goto bad; for (i = 0; i < sp->timecnt; ++i) { sp->ats[i] = (stored == 4) ? detzcode(p) : detzcode64(p); @@ -421,7 +426,7 @@ for (i = 0; i < sp->timecnt; ++i) { sp->types[i] = (unsigned char) *p++; if (sp->types[i] >= sp->typecnt) - return -1; + goto bad; } for (i = 0; i < sp->typecnt; ++i) { register struct ttinfo * ttisp; @@ -431,11 +436,11 @@ p += 4; ttisp->tt_isdst = (unsigned char) *p++; if (ttisp->tt_isdst != 0 && ttisp->tt_isdst != 1) - return -1; + goto bad; ttisp->tt_abbrind = (unsigned char) *p++; if (ttisp->tt_abbrind < 0 || ttisp->tt_abbrind > sp->charcnt) - return -1; + goto bad; } for (i = 0; i < sp->charcnt; ++i) sp->chars[i] = *p++; @@ -460,7 +465,7 @@ ttisp->tt_ttisstd = *p++; if (ttisp->tt_ttisstd != TRUE && ttisp->tt_ttisstd != FALSE) - return -1; + goto bad; } } for (i = 0; i < sp->typecnt; ++i) { @@ -473,7 +478,7 @@ ttisp->tt_ttisgmt = *p++; if (ttisp->tt_ttisgmt != TRUE && ttisp->tt_ttisgmt != FALSE) - return -1; + goto bad; } } /* @@ -506,11 +511,11 @@ /* ** If this is an old file, we're done. */ - if (u.tzhead.tzh_version[0] == '\0') + if (u->tzhead.tzh_version[0] == '\0') break; - nread -= p - u.buf; + nread -= p - u->buf; for (i = 0; i < nread; ++i) - u.buf[i] = p[i]; + u->buf[i] = p[i]; /* ** If this is a narrow integer time_t system, we're done. */ @@ -518,13 +523,13 @@ break; } if (doextend && nread > 2 && - u.buf[0] == '\n' && u.buf[nread - 1] == '\n' && + u->buf[0] == '\n' && u->buf[nread - 1] == '\n' && sp->typecnt + 2 <= TZ_MAX_TYPES) { struct state ts; register int result; - u.buf[nread - 1] = '\0'; - result = tzparse(&u.buf[1], &ts, FALSE); + u->buf[nread - 1] = '\0'; + result = tzparse(&u->buf[1], &ts, FALSE); if (result == 0 && ts.typecnt == 2 && sp->charcnt + ts.charcnt <= TZ_MAX_CHARS) { for (i = 0; i < 2; ++i) @@ -568,7 +573,11 @@ break; } } + free(u); return 0; +bad: + free(u); + return -1; } static int

1 0

2010m
by Olson, Arthur David (NIH/NCI) [E] Nov. 21, 2010

Nov. 21, 2010

The files... ftp://elsie.nci.nih.gov/pub/tzcode2010m.tar.gz ...and... ftp://elsie.nci.nih.gov/pub/tzdata2010m.tar.gz ...are now available; these reflect the Hong Kong, Vostok, and zic.c changes circulated recently on the time zone mailing list. --ado

3 2

DST in the news :-)
by Olson, Arthur David (NIH/NCI) [E] Nov. 10, 2010

Nov. 10, 2010

"I put myself and my staff through this crazy, huge ordeal, all because I refused to go on at midnight, okay? And so I work, you know, and then I get this job at eleven, supposed to be a big deal. Then yesterday daylight savings [sic] time ended. Right now it's basically midnight." Conan O'Brien on the 2010-11-08 premier of "Conan."

1 0

FW: zone.tab change request
by Olson, Arthur David (NIH/NCI) [E] Nov. 8, 2010

Nov. 8, 2010

I'm forwarding this message from an address that's not on the time zone mailing list. Those of you who are on the list, please direct replies appropriately. --ado From: blacklight332(a)aol.com [mailto:blacklight332@aol.com] Sent: Friday, November 05, 2010 7:10 To: tz(a)lecserver.nci.nih.gov Subject: zone.tab change request tz(a)elsie.nci.nih.gov can you change the zone.tab file to say: ID -0610+10648 Asia/Jakarta Java & Sumatra ID -0002+10920 Asia/Pontianak west & central Borneo(Kalimantan) ID -0507+11924 Asia/Makassar east & south Borneo, Celebes(Sulawesi), Bali, Nusa Tengarra, west Timor ID -0232+14042 Asia/Jayapura Irian Jaya (Papua) & the Moluccas (Maluku) instead of ID -0610+10648 Asia/Jakarta Java & Sumatra ID -0002+10920 Asia/Pontianak west & central Borneo ID -0507+11924 Asia/Makassar east & south Borneo, Celebes, Bali, Nusa Tengarra, west Timor ID -0232+14042 Asia/Jayapura Irian Jaya & the Moluccas the regions have changed their names in 2007. bl

1 0

zone.tab change request
by blacklight332＠aol.com Nov. 5, 2010

Nov. 5, 2010

tz(a)elsie.nci.nih.gov can you change the zone.tab file to say: ID -0610+10648 Asia/Jakarta Java & Sumatra ID -0002+10920 Asia/Pontianak west & central Borneo(Kalimantan) ID -0507+11924 Asia/Makassar east & south Borneo, Celebes(Sulawesi), Bali, Nusa Tengarra, west Timor ID -0232+14042 Asia/Jayapura Irian Jaya (Papua) & the Moluccas (Maluku) instead of ID -0610+10648 Asia/Jakarta Java & Sumatra ID -0002+10920 Asia/Pontianak west & central Borneo ID -0507+11924 Asia/Makassar east & south Borneo, Celebes, Bali, Nusa Tengarra, west Timor ID -0232+14042 Asia/Jayapura Irian Jaya & the Moluccas the regions have changed their names in 2007. bl

1 0

Acre (Brazil) would like to return to old time zone GMT-5
by Alexander Krivenyshev Nov. 1, 2010

Nov. 1, 2010

According to the media from Brazil, during recent referendum state Acre rejected current time zone (GMT-4) and would like to return to old time zone (pre-2008) - GMT-5. The decision of Acre will be sent to the Congress to become law and to re-apply the old time. (in Portuguese): Acreanos decidem pelo horário antigo http://www.noticiasdahora.com/acre/4184-acreanos-decidem-voltar-ao-horario- antig.html Acreanos decidem voltar ao horário antigo em votação do referendo http://portalamazonia.globo.com/pscript/noticias/noticias.php?pag=old&idN=1… or http://www.worldtimezone.com/dst_news/dst_news_brazil07.html Quick reference page for current (2010) and old (pre-2008) Brazil Time Zones map http://www.worldtimezone.com/brazil-time-new-old.php Alexander Krivenyshev, http://www.worldtimezone.com

1 0

data2010o
by Olson, Arthur David (NIH/NCI) [E] Nov. 1, 2010

Nov. 1, 2010

The file... ftp://elsie.nci.nih.gov/pub/tzdata2010o.tar.gz ...is now available; this includes the change to the end of DST in Fiji in 2011 circulated last week on the time zone mailing list. --ado

1 0

proposed time zone package change (Fiji's 2001 DST end)
by Arthur David Olson Oct. 28, 2010

Oct. 28, 2010

Below is a proposed change to the time zone package's "australasia" file, moving Fiji's 2011 end of DST from March 27 to March 6. If there are no problems, the ftp site will be updated on 2010-11-01. --ado ------- australasia ------- *** /tmp/geta22811 Tue Oct 26 17:01:05 2010 --- /tmp/getb22811 Tue Oct 26 17:01:05 2010 *************** *** 1,5 **** # <pre> ! # @(#)australasia 8.19 # This file is in the public domain, so clarified as of # 2009-05-17 by Arthur David Olson. --- 1,5 ---- # <pre> ! # @(#)australasia 8.20 # This file is in the public domain, so clarified as of # 2009-05-17 by Arthur David Olson. *************** *** 284,289 **** --- 284,302 ---- # http://www.timeanddate.com/news/time/fiji-dst-ends-march-2010.html # </a> + # From Alexander Krivenyshev (2010-10-24): + # According to Radio Fiji and Fiji Times online, Fiji will end DST 3 + # weeks earlier than expected - on March 6, 2011, not March 27, 2011... + # Here is confirmation from Government of the Republic of the Fiji Islands, + # Ministry of Information (fiji.gov.fj) web site: + # <a href="http://www.fiji.gov.fj/index.php?option=com_content&view=article&id=2608:da…"> + # http://www.fiji.gov.fj/index.php?option=com_content&view=article&id=2608:da… + # </a> + # or + # <a href="http://www.worldtimezone.com/dst_news/dst_news_fiji04.html"> + # http://www.worldtimezone.com/dst_news/dst_news_fiji04.html + # </a> + # Rule NAME FROM TO TYPE IN ON AT SAVE LETTER/S Rule Fiji 1998 1999 - Nov Sun>=1 2:00 1:00 S Rule Fiji 1999 2000 - Feb lastSun 3:00 0 - *************** *** 290,296 **** Rule Fiji 2009 only - Nov 29 2:00 1:00 S Rule Fiji 2010 only - Mar lastSun 3:00 0 - Rule Fiji 2010 only - Oct 24 2:00 1:00 S ! Rule Fiji 2011 only - Mar lastSun 3:00 0 - # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Pacific/Fiji 11:53:40 - LMT 1915 Oct 26 # Suva 12:00 Fiji FJ%sT # Fiji Time --- 303,309 ---- Rule Fiji 2009 only - Nov 29 2:00 1:00 S Rule Fiji 2010 only - Mar lastSun 3:00 0 - Rule Fiji 2010 only - Oct 24 2:00 1:00 S ! Rule Fiji 2011 only - Mar Sun>=1 3:00 0 - # Zone NAME GMTOFF RULES FORMAT [UNTIL] Zone Pacific/Fiji 11:53:40 - LMT 1915 Oct 26 # Suva 12:00 Fiji FJ%sT # Fiji Time

2 2

Extension to tzcode to support additional timezones
by John Baldwin Oct. 27, 2010

Oct. 27, 2010

Several of my employers have needed to work with data from multiple timezones within a single process. The traditional way to work with this has been to change the value of the TZ environment variable and invoke tzset() and switch between timezones one at a time. What I have done is extend the time(3) API to allow applications to load and use arbitrary timezones separate from the current local and GM timezones. I've done this by adding the following API calls: - void *tzopen(const char *name). This loads the rules for a specified timezone and returns a void * cookie. If the zone cannot be parsed it returns NULL and sets errno to EINVAL. - void tzclose(void *zone). This "closes" a set of rules opened via tzopen() by releasing the associated resources. - struct tm *tztime(void *zone, const time_t *t, struct tm *tm). This is like localtime_r() except that it uses the timezone rules from the passed in cookie instead of the local timezone indicated by TZ. - time_t timetz(void *zone, struct tm * const tm). This is like mktime() except that it uses the timezone rules from the passed in cookie instead of the local timezone indicated by TZ. The changes to the code are fairly simple since the per-zone state was already broken out into a standalone state object to handle GMT vs local time. In some cases the patches make things cleaner (I think) by removing hacks where a function pointer was queried to determine which state object to use since the state object pointer is now passed all the way down the internal time code stack. I have a patch that is relative to the FreeBSD source tree, but all of the real code changes are in localtime.c which I believe should apply directly to the tzcode distribution. The patch is available at http://www.FreeBSD.org/~jhb/patches/tzopen.patch Is this API something that you folks would be interested in? I currently have this as a private patch locally, but if possible I would like it adopted upstream. -- John Baldwin

9 22