From owner-freebsd-current@FreeBSD.ORG Mon May 18 19:06:01 2009 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id C37D11065674; Mon, 18 May 2009 19:06:01 +0000 (UTC) (envelope-from jhb@freebsd.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 934AF8FC28; Mon, 18 May 2009 19:06:01 +0000 (UTC) (envelope-from jhb@freebsd.org) Received: from bigwig.baldwin.cx (66.111.2.69.static.nyinternet.net [66.111.2.69]) by cyrus.watson.org (Postfix) with ESMTPSA id 422A946B0C; Mon, 18 May 2009 15:06:01 -0400 (EDT) Received: from jhbbsd.hudson-trading.com (unknown [209.249.190.8]) by bigwig.baldwin.cx (Postfix) with ESMTPA id A38008A026; Mon, 18 May 2009 15:05:59 -0400 (EDT) From: John Baldwin To: Attilio Rao Date: Mon, 18 May 2009 14:57:24 -0400 User-Agent: KMail/1.9.7 References: <08D7DC2A-68BE-47B6-8D5D-5DE6B48F87E5@wanderview.com> <200905181331.11174.jhb@freebsd.org> <3bbf2fe10905181038geaec26csffea4788a40feaca@mail.gmail.com> In-Reply-To: <3bbf2fe10905181038geaec26csffea4788a40feaca@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Message-Id: <200905181457.25238.jhb@freebsd.org> X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.0.1 (bigwig.baldwin.cx); Mon, 18 May 2009 15:06:00 -0400 (EDT) X-Virus-Scanned: clamav-milter 0.95 at bigwig.baldwin.cx X-Virus-Status: Clean X-Spam-Status: No, score=-2.5 required=4.2 tests=AWL,BAYES_00,RDNS_NONE autolearn=no version=3.2.5 X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on bigwig.baldwin.cx Cc: Adam McDougall , freebsd-current@freebsd.org, Artem Belevich , Ben Kelly Subject: Re: [patch] zfs livelock and thread priorities X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 18 May 2009 19:06:02 -0000 On Monday 18 May 2009 1:38:03 pm Attilio Rao wrote: > 2009/5/18 John Baldwin : > > On Monday 18 May 2009 1:12:59 pm Attilio Rao wrote: > >> 2009/5/18 John Baldwin : > >> > On Saturday 16 May 2009 12:40:44 pm Ben Kelly wrote: > >> 2) I think this KPI can be dangerous and lead to problems. Priority is > >> something highly fragile. > > > > All the more reason to make developers _think_ about the priority of ea= ch > > kthread they create. =C2=A0Right now all these threads start out with a= priority > > of PVM since that is what thread0 runs at. =C2=A0Does that sound right = to you? =C2=A0Do > > you think many folks realize that? =C2=A0It sounds very bogus to me. = =C2=A0I think > > forcing people to pick a sensible priority for each thread is far bette= r than > > the complete lack of thought that often happens now. >=20 > At least, we could leave the default version not accepting any > priority for threads which are not interested on that and trying to > move people to the new KPI _only and if only_ they need real boosts or > lay down. I would rather force people to think. We've had problems in the past with folks not thinking clearly enough (e.g. just using a constant to tsleep() instead of figuring out a real timeout value to use). =2D-=20 John Baldwin