From owner-freebsd-current@FreeBSD.ORG Wed Sep 1 21:14:08 2004 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9EC5816A65D for ; Wed, 1 Sep 2004 21:14:08 +0000 (GMT) Received: from mail2.speakeasy.net (mail2.speakeasy.net [216.254.0.202]) by mx1.FreeBSD.org (Postfix) with ESMTP id 79E7143D48 for ; Wed, 1 Sep 2004 21:14:08 +0000 (GMT) (envelope-from jhb@FreeBSD.org) Received: (qmail 7049 invoked from network); 1 Sep 2004 21:14:08 -0000 Received: from dsl027-160-063.atl1.dsl.speakeasy.net (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) encrypted SMTP for ; 1 Sep 2004 21:14:07 -0000 Received: from [10.50.41.228] (gw1.twc.weather.com [216.133.140.1]) (authenticated bits=0) by server.baldwin.cx (8.12.11/8.12.11) with ESMTP id i81LDVpI062794; Wed, 1 Sep 2004 17:14:01 -0400 (EDT) (envelope-from jhb@FreeBSD.org) From: John Baldwin To: freebsd-current@FreeBSD.org Date: Wed, 1 Sep 2004 16:36:22 -0400 User-Agent: KMail/1.6.2 References: <41303747.4050104@elischer.org> In-Reply-To: <41303747.4050104@elischer.org> MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <200409011636.22089.jhb@FreeBSD.org> X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on server.baldwin.cx cc: Julian Elischer cc: FreeBSD current mailing list Subject: Re: hlt_cpus_mask X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 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: Wed, 01 Sep 2004 21:14:08 -0000 On Saturday 28 August 2004 03:41 am, Julian Elischer wrote: > This is defined static in i386 and amd64, > but as it is not exported it can not be used in > subr_smp.c, which probably should mask against it when trying to > forward roundrobin and signals. > > should this be a globally available thing? > Otherwise we are sending IPIs to threads that are > not capable of doing anything with them. More correct might be to add an 'active_cpus' mask that contains the set of active CPUs and use that in place of all_cpus where appropriate. active_cpus would basically be 'all_cpus & ~hlt_cpus_mask', but it probably makes more sense from an MI perspective since 'hlt' is fairly x86-specific. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve" = http://www.FreeBSD.org