This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug time/19738] __tzfile_default 32-bit time_t overflow
- From: "carlos at redhat dot com" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sourceware dot org
- Date: Thu, 10 Mar 2016 01:56:21 +0000
- Subject: [Bug time/19738] __tzfile_default 32-bit time_t overflow
- Auto-submitted: auto-generated
- References: <bug-19738-131 at http dot sourceware dot org/bugzilla/>
https://sourceware.org/bugzilla/show_bug.cgi?id=19738
--- Comment #15 from Carlos O'Donell <carlos at redhat dot com> ---
(In reply to Paul Eggert from comment #14)
> (In reply to Carlos O'Donell from comment #13)
> > So in summary:
> > * Only impacts 32-bit targets
> > * For users which have custom TZ values that don't map to exiting rules
> > files.
> > * Requires the use of a new posixrules compiled with a zic that includes
> > fixes present in tzcode 2014c (2014-05-13) and later.
>
> While that's all correct, it's a bit scarier than reality. Perhaps you could
> append to the 2nd bullet the phrase "and that specify daylight saving time
> without giving daylight saving rules". E.g., as sbr mentions, the bug occurs
> with TZ='PST+8PDT' but does not occur with TZ='PST+8PDT,M3.2.0,M11.1.1'.
OK, how about this:
(1) Only impacts 32-bit targets.
(2) For users which have custom TZ values that don't map to
exiting zone files (triggers the use of default posix rules)
e.g. PST+8PDT instead of PST8PDT (which has a zone file)
and that specify daylight saving time e.g. PDT, without
giving daylight saving rules e.g. TZ='PST+8PDT' triggers the
bug, but TZ='PST+8PDT,M3.2.0,M11.1.1' does not (has daylight
savings rule explicitly encoded).
(3) Requires the use of a new posixrules compiled with a zic
that includes fixes present in tzcode 2014c (2014-05-13) and
later.
The fix to zic.c has been in upstream since 2014-06-25, so it impacts 2.20 and
newer releases.
This impacts Fedora 21 (Unsupported), 22, 23, 24 (Alpha), and Rawhide. However,
given that nobody has ever reported this is probably an indicator of how rare
the usage is in practice.
So I think we can just fix this in master and declare it done.
--
You are receiving this mail because:
You are on the CC list for the bug.