From owner-cvs-all Tue Dec 5 15:21: 1 2000 From owner-cvs-all@FreeBSD.ORG Tue Dec 5 15:20:56 2000 Return-Path: Delivered-To: cvs-all@freebsd.org Received: from fw.wintelcom.net (ns1.wintelcom.net [209.1.153.20]) by hub.freebsd.org (Postfix) with ESMTP id AA45E37B400; Tue, 5 Dec 2000 15:20:56 -0800 (PST) Received: (from bright@localhost) by fw.wintelcom.net (8.10.0/8.10.0) id eB5NKtn03673; Tue, 5 Dec 2000 15:20:55 -0800 (PST) Date: Tue, 5 Dec 2000 15:20:55 -0800 From: Alfred Perlstein To: Daniel Eischen Cc: "David O'Brien" , cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/vm phys_pager.c Message-ID: <20001205152054.M8051@fw.wintelcom.net> References: <20001205145908.K8051@fw.wintelcom.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from eischen@vigrid.com on Tue, Dec 05, 2000 at 06:11:14PM -0500 Sender: bright@fw.wintelcom.net Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG * Daniel Eischen [001205 15:12] wrote: > On Tue, 5 Dec 2000, Alfred Perlstein wrote: > > * David O'Brien [001205 14:31] wrote: > > > On Tue, Dec 05, 2000 at 02:23:11PM -0800, Alfred Perlstein wrote: > > > > Modified files: (Branch: RELENG_4) > > > > sys/vm phys_pager.c > > > > Log: > > > > MFC: 1.4 (handle non PAGE_SIZE requests properly) > > > > > > Uh.. what happened to letting it settle in -current first... > > > Didn't this *same* discussion come up just last week... perhaps you are > > > a little too 'D'elete key happy. > > > > 1) I tested it under -stable before -current. > > 2) I needed it working. > > 3) It was a bugfix and needed. > > 4) It defaults to off in all systems. > > 5) It was reviewed. > > 6) I have extreme doubts that anyone could have been using it reliably > > before my fix. > > 7) It's been broken for many months, I finally spent all night tracking > > it down and the Maintainer (peter) didn't seem interested in chasing > > down the problem when I told him about it months ago. > > I don't want to pick on Alfred; just using this as a platform. > > For problems that don't affect the majority of folks and is not > a security related problem, there is no reason *not* to wait to > MFC. I'd like to see all fixes like this settle for at least > a day or two, just to make sure that world wasn't accidentally > broken. And who exactly would bump into this problem except me? It was a one line change, an 'off by PAGE_SIZE' error and no one was going to test it in -current anyhow. This code has been broken since it was brought into the tree, no one is using it otherwise we'd hear complaints. This is completely the wrong place to start this arguement, in fact it's getting so annoying that I've been thinking of keeping my own private branch of -stable to avoid the _real_ premature commits that mess me up and keep my fixes to myself. Maybe I'll get around to allowing access to it, maybe not. *feh* .. whatever... You'll note the that loadable syscipc stuff hasn't been MFC'd because I'm still hashing out bugs with it and I knew it could be problematic. I think a bit more faith in my work would be nice. -- -Alfred Perlstein - [bright@wintelcom.net|alfred@freebsd.org] "I have the heart of a child; I keep it in a jar on my desk." To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message