From owner-freebsd-current Fri Dec 11 00:48:36 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA06696 for freebsd-current-outgoing; Fri, 11 Dec 1998 00:48:36 -0800 (PST) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from bright.fx.genx.net (bright.fx.genx.net [206.64.4.154]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA06691 for ; Fri, 11 Dec 1998 00:48:35 -0800 (PST) (envelope-from bright@hotjobs.com) Received: from localhost (bright@localhost) by bright.fx.genx.net (8.9.1/8.9.1) with ESMTP id DAA01216; Fri, 11 Dec 1998 03:52:27 -0500 (EST) (envelope-from bright@hotjobs.com) X-Authentication-Warning: bright.fx.genx.net: bright owned process doing -bs Date: Fri, 11 Dec 1998 03:52:27 -0500 (EST) From: Alfred Perlstein X-Sender: bright@bright.fx.genx.net To: Warner Losh cc: Matthew Dillon , Archie Cobbs , jwd@unx.sas.com, freebsd-current@FreeBSD.ORG Subject: Re: inetd: realloc/free bug In-Reply-To: <199812110722.AAA00650@harmony.village.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Fri, 11 Dec 1998, Warner Losh wrote: > In message <199812110717.XAA35358@apollo.backplane.com> Matthew Dillon writes: > : I seem to recall linux returning the time remaining a long time ago, > : but I think they ripped it out. I don't have a linux box handy to > : test the linux current kernel. > > Linux implemented the timeval stomping version of select. It was the > first OS to do this. At the time they claimed that BSD 4.4 would be > doing this and that they were going to be compatible with that, plus > it was listed as a bug. BSD 4.4 came out, and it didn't change this > part of the interface. There were boatloads of programs that were > inexpertly ported to Linux that exhibited cpu eating problems. Turns > out that they weren't robust enough to deal with the timeval stomping > interface. In time they created a bsd_select, which was what all user > progams used. I think it is possible to get the "new" linux behavior, > but I don't know how. > > : I definitely think it's a mistake to change select()'s timeout > : operation, too many programs pre-set the timeval and assume it will > : not be changed on multiple calls to select(). If we were to > : implement a returned-time it would have to be with a new system > : call. > > Agreed. ugh, all my wasted cycles resetting the timeval "just in case" select2() anyone? Alfred Perlstein - Programmer, HotJobs Inc. - www.hotjobs.com -- There are operating systems, and then there's FreeBSD. -- http://www.freebsd.org/ 3.0-current > Warner > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-current" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message