Re: svlogd timestamps fail on OpenBSD 5.5 (snapshot)

From: Jérémie Courrèges-Anglas <jca_at_wxcvbn.org>
Date: Tue, 22 Apr 2014 13:02:54 +0200

Mike Jackson <mj_at_netauth.com> writes:

> On 22 Apr 2014, at 02.52, Jérémie Courrèges-Anglas <jca_at_wxcvbn.org> wrote:
>
>> jca_at_wxcvbn.org (Jérémie Courrèges-Anglas) writes:
>>
>> You funny ezmlm, where have you hidden my patch?
>>
>> This is not a blind guess anymore, please confirm that it fix your
>> problem.
>>
>> --- src/fmt_ptime.c.orig Tue Nov 21 16:14:12 2006
>> +++ src/fmt_ptime.c Tue Apr 22 01:27:37 2014
>> _at_@ -4,7 +4,7 @@
>>
>> unsigned int fmt_ptime2(char *s, struct taia *ta, char sep) {
>> struct tm *t;
>> - unsigned long u;
>> + time_t u;
>>
>> if (ta->sec.x < 4611686018427387914ULL) return(0); /* impossible? */
>> u =ta->sec.x -4611686018427387914ULL;
>>
>
>
> Hi,
>
> Yes, the machine in question has a very old 32-bit epia board.
>
> Problem solved. Thanks.

Fixed in the -current OpenBSD ports tree.

> Might be time for a new release that can handle the move to 64-bit time systems. I’m guessing that most of them will be making the change soon.

Only NetBSD and OpenBSD have done so, AFAIK. Some others use a "long"
time_t and thus don't attempt to fix the problem for 32 bits archs
(except the x32 Linux ABI, IIUC). Given the discussions I've read on
other OSes mailing lists, I doubt this will change soon.

Thanks,
-- 
jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF  DDCC 0DFA 74AE 1524 E7EE
Received on Tue Apr 22 2014 - 11:02:54 UTC

This archive was generated by hypermail 2.3.0 : Sun May 09 2021 - 19:44:18 UTC