From owner-freebsd-rc@FreeBSD.ORG Tue Oct 2 00:46:30 2007 Return-Path: Delivered-To: freebsd-rc@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 7A53E16A468 for ; Tue, 2 Oct 2007 00:46:30 +0000 (UTC) (envelope-from dougb@FreeBSD.org) Received: from mail2.fluidhosting.com (mx21.fluidhosting.com [204.14.89.4]) by mx1.freebsd.org (Postfix) with SMTP id 43DB413C457 for ; Tue, 2 Oct 2007 00:46:30 +0000 (UTC) (envelope-from dougb@FreeBSD.org) Received: (qmail 17439 invoked by uid 399); 2 Oct 2007 00:46:28 -0000 Received: from localhost (HELO lap.dougb.net) (dougb@dougbarton.us@127.0.0.1) by localhost with ESMTP; 2 Oct 2007 00:46:28 -0000 X-Originating-IP: 127.0.0.1 Date: Mon, 1 Oct 2007 17:46:25 -0700 (PDT) From: Doug Barton To: Mike Makonnen In-Reply-To: <20071001224116.GA82760@terra.mike.lan> Message-ID: References: <584bfc3f0709300300s22f2606w3f2628edc1aa15f@mail.gmail.com> <20071001224116.GA82760@terra.mike.lan> X-message-flag: Outlook -- Not just for spreading viruses anymore! X-OpenPGP-Key-ID: 0xD5B2F0FB Organization: http://www.FreeBSD.org/ MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: freebsd-rc@freebsd.org Subject: Re: rc.d cleanup patch redux X-BeenThere: freebsd-rc@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: "Discussion related to /etc/rc.d design and implementation." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 02 Oct 2007 00:46:30 -0000 On Tue, 2 Oct 2007, Mike Makonnen wrote: > It's just that this little chage adds a little extra processig time here, > that change adds just a little extra there, etc... and before you know it > they all add up to a big difference (especially if we don't use faststop > like des suggested). I feel the same way, and I am resistant to ideas that add "just a little bit" of time for the same reason, especially for boot stuff. However with this change I literally could not measure any difference. The shutdown time for my system ranged from about 4 seconds to about 8 seconds, whether the patch was applied or not. Most of the time attributed to shutting down daemons actually comes from the ports I have installed. If someone had a very large number of services from the base running this change may introduce a measurable change in shutdown time, but I really doubt it. > I'm not really against this change, I just wanted to voice my reservations. > In fact, I would be more comfortable with leaving this last part of > the change until after the code-freeze since it's also likely to > introduce more "foo is not running?" console spammage. Yeah, my original patch did not include the shutdown stuff, I added it in response to a thread elsewhere that basically amounted to "lots of other rc.d scripts don't have it, so I shouldn't have to add it either." I'll split the patch into "functional" and "non-functional" halves, ask for re@ approval for the latter, and post the former to -current to ask for more testers. Thanks, Doug -- This .signature sanitized for your protection