Horton

Re: gettimeofday() goes backwards on FreeBSD 4.9

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: "Nigel J. Andrews" <nandrews@investsystems.co.uk>
Cc: Darcy Buskermolen <darcy@wavefire.com>, "Marc G. Fournier" <scrappy@postgresql.org>, pgsql-hackers@postgresql.org
Date: 2003-11-29T01:08:31Z
Lists: pgsql-hackers 
"Nigel J. Andrews" <nandrews@investsystems.co.uk> writes:
> On Fri, 28 Nov 2003, Tom Lane wrote:
>> "Nigel J. Andrews" <nandrews@investsystems.co.uk> writes:
> On an Intel Linux 2.4.18 I get them quite often, 25 in 1'45", but they
> are all just a microsecond.
>> 
>> What do you mean by "just a microsecond"?

> I mean it's always a "out of order tv_usec..." line and the difference is
> 1us. That is a.out gives:

> out of order tv_usec: 1070065862 374978, prev 1070065862 374979
> out of order tv_usec: 1070065867 814300, prev 1070065867 814301

Fascinating.  I'd call that a bug too, but evidently one with a
different mechanism than the BSD issue we are chasing.

FWIW, I have not seen any failures in a fair amount of runtime on
a 2.4.18 (Red Hat 8.0) kernel here, running on a Dell P4.  What is
the hardware platform you're using?

			regards, tom lane