Date: Fri, 19 Jul 2002 08:06:05 +0100 From: David Malone <dwmalone@maths.tcd.ie> To: Jeremy D'Hoinne <jeremy.dhoinne@netasq.com> Cc: freebsd-hackers@FreeBSD.org Subject: Re: select() behavior when system date changes precisions. Message-ID: <20020719070605.GA60918@walton.maths.tcd.ie> In-Reply-To: <20020718175856.2a5dd923.jeremy.dhoinne@netasq.com> References: <20020718175856.2a5dd923.jeremy.dhoinne@netasq.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jul 18, 2002 at 05:58:56PM +0200, Jeremy D'Hoinne wrote: > I've forgottent to precise that servers involved use thread library > (compilation with -pthread flag) Ahhh - this may explain what you are seeing - I think that the threaded library uses gettimeofday, which would see the time going backwards. The non-threaded version of select uses a kernel "getmicrouptime" call, which shouldn't have this problem. It is possible a getuptime call might be useful for people implimenting things like the threading library. OTOH, there is usually no need to step the clock backwards on FreeBSD after boot. Because get gettimeofday returns the time in UTC there will be no jump for daylight savings. The only remaining problem are leap seconds, which don't happen very often. David. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020719070605.GA60918>