From owner-freebsd-alpha Sun Sep 1 2:26:42 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2313637B400; Sun, 1 Sep 2002 02:26:38 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1E2A843E3B; Sun, 1 Sep 2002 02:26:37 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g819QaJR003816; Sun, 1 Sep 2002 11:26:36 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g819QZF9003815; Sun, 1 Sep 2002 11:26:35 +0200 (CEST) Date: Sun, 1 Sep 2002 11:26:35 +0200 From: Wilko Bulte To: Marco Beishuizen Cc: FreeBSD alpha mailing list , FreeBSD questions mailing list Subject: Re: disk hardware failure? Message-ID: <20020901112635.C3683@freebie.xs4all.nl> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from mbeis@wxs.nl on Sun, Sep 01, 2002 at 04:02:31AM +0200 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Sun, Sep 01, 2002 at 04:02:31AM +0200, Marco Beishuizen wrote: > > Hi all, > > >From one moment to another the second harddisk of my PWS 600a fails to > operate. It looks very strange to me because the disk isn't used at > all. I suddenly got the following messages lots of times: > > /kernel: (da1: isp0:0:1:0): > READ CAPACITY. CDB: 25 0 0 0 0 0 0 0 0 > HARDWARE FAILURE asc:9,ee > Vendor Specific ASCQ field > replacable unit: 80 sks: 80,0 > > Disk da0 is my boot disk where FreeBSD is installed and works fine. > Disk da1 is only used to store files, but didn't have any files on it > at this time. > Disk da1 is mounted at startup (line in /etc/fstab) in /disk_da1. > > Now I can't boot FreeBSD because fsck finds errors on da1. I looked > inside and no cables are loose or anything. > > Is my disk really screwed up just like that, or could it be something > else? The disk is really screwed, generally SCSI messages like this tend to be accurate (unfortunately in this case) -- | / o / /_ _ |/|/ / / /( (_) Bulte wilko@FreeBSD.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sun Sep 1 3: 5: 3 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 927E337B400; Sun, 1 Sep 2002 03:04:53 -0700 (PDT) Received: from hotmail.com (oe15.pav2.hotmail.com [64.4.36.119]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4B92643E42; Sun, 1 Sep 2002 03:04:53 -0700 (PDT) (envelope-from unixtools@hotmail.com) Received: from mail pickup service by hotmail.com with Microsoft SMTPSVC; Sun, 1 Sep 2002 03:04:53 -0700 X-Originating-IP: [202.41.224.36] From: "Unix Tools" To: "Marco Beishuizen" , "FreeBSD alpha mailing list" , "FreeBSD questions mailing list" References: Subject: Re: disk hardware failure? Date: Sun, 1 Sep 2002 03:49:32 +0530 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2600.0000 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000 Message-ID: X-OriginalArrivalTime: 01 Sep 2002 10:04:53.0214 (UTC) FILETIME=[0436B7E0:01C2519F] Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org drop to single user mode and then run fsck because if you are not in single user mode the changes made to the filesystem will be considered as errors by the fsck. ----- Original Message ----- From: "Marco Beishuizen" To: "FreeBSD alpha mailing list" ; "FreeBSD questions mailing list" Sent: Sunday, September 01, 2002 07:32 AM Subject: disk hardware failure? > > Hi all, > > From one moment to another the second harddisk of my PWS 600a fails to > operate. It looks very strange to me because the disk isn't used at > all. I suddenly got the following messages lots of times: > > /kernel: (da1: isp0:0:1:0): > READ CAPACITY. CDB: 25 0 0 0 0 0 0 0 0 > HARDWARE FAILURE asc:9,ee > Vendor Specific ASCQ field > replacable unit: 80 sks: 80,0 > > Disk da0 is my boot disk where FreeBSD is installed and works fine. > Disk da1 is only used to store files, but didn't have any files on it > at this time. > Disk da1 is mounted at startup (line in /etc/fstab) in /disk_da1. > > Now I can't boot FreeBSD because fsck finds errors on da1. I looked > inside and no cables are loose or anything. > > Is my disk really screwed up just like that, or could it be something > else? > > Thanks in advance, > > Marco > > > -- > The only problem with being a man of leisure is that you can never stop > and take a rest. > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-questions" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sun Sep 1 7:11:38 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 62DDA37B400; Sun, 1 Sep 2002 07:11:32 -0700 (PDT) Received: from smtp01.wxs.nl (smtp01.wxs.nl [195.121.6.61]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2BA3543E6A; Sun, 1 Sep 2002 07:11:31 -0700 (PDT) (envelope-from mbeis@wxs.nl) Received: from yokozuna.bsd ([213.10.35.60]) by smtp01.wxs.nl (Netscape Messaging Server 4.15) with ESMTP id H1RJF501.9D8; Sun, 1 Sep 2002 16:11:29 +0200 Date: Sun, 1 Sep 2002 16:11:28 +0200 (CEST) From: Marco Beishuizen Reply-To: Marco Beishuizen To: Wilko Bulte Cc: FreeBSD alpha mailing list , FreeBSD questions mailing list Subject: Re: disk hardware failure? => new disk In-Reply-To: <20020901112635.C3683@freebie.xs4all.nl> Message-ID: Homepage: FreeBSD: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Sun, 1 Sep 2002, the wise Wilko Bulte spoke, and said: > On Sun, Sep 01, 2002 at 04:02:31AM +0200, Marco Beishuizen wrote: > > > > Hi all, > > > > >From one moment to another the second harddisk of my PWS 600a fails to > > operate. It looks very strange to me because the disk isn't used at > > all. I suddenly got the following messages lots of times: > > > > /kernel: (da1: isp0:0:1:0): > > READ CAPACITY. CDB: 25 0 0 0 0 0 0 0 0 > > HARDWARE FAILURE asc:9,ee > > Vendor Specific ASCQ field > > replacable unit: 80 sks: 80,0 > > > > Disk da0 is my boot disk where FreeBSD is installed and works fine. > > Disk da1 is only used to store files, but didn't have any files on it > > at this time. > > Disk da1 is mounted at startup (line in /etc/fstab) in /disk_da1. > > > > Now I can't boot FreeBSD because fsck finds errors on da1. I looked > > inside and no cables are loose or anything. > > > > Is my disk really screwed up just like that, or could it be something > > else? > > The disk is really screwed, generally SCSI messages like this tend > to be accurate (unfortunately in this case) > > Well, the disk seems really broken. It's not accessable anymore. Is it possible to put a new, modern SCSI-disk in the Alpha? Or is the SCSI-connection changed. Marco -- All other things being equal, a bald man cannot be elected President of the United States. -- Vic Gold To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sun Sep 1 7:41:30 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id A43B337B400; Sun, 1 Sep 2002 07:41:21 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7C07C43E6A; Sun, 1 Sep 2002 07:41:20 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g81EfJJR004446; Sun, 1 Sep 2002 16:41:19 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g81EfIw9004445; Sun, 1 Sep 2002 16:41:18 +0200 (CEST) Date: Sun, 1 Sep 2002 16:41:18 +0200 From: Wilko Bulte To: Marco Beishuizen Cc: FreeBSD alpha mailing list , FreeBSD questions mailing list Subject: Re: disk hardware failure? => new disk Message-ID: <20020901164118.A4222@freebie.xs4all.nl> References: <20020901112635.C3683@freebie.xs4all.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from mbeis@wxs.nl on Sun, Sep 01, 2002 at 04:11:28PM +0200 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Sun, Sep 01, 2002 at 04:11:28PM +0200, Marco Beishuizen wrote: > On Sun, 1 Sep 2002, the wise Wilko Bulte spoke, and said: > > > On Sun, Sep 01, 2002 at 04:02:31AM +0200, Marco Beishuizen wrote: > > > > > > Hi all, > > > > > > >From one moment to another the second harddisk of my PWS 600a fails to > > > operate. It looks very strange to me because the disk isn't used at > > > all. I suddenly got the following messages lots of times: > > > > > > /kernel: (da1: isp0:0:1:0): > > > READ CAPACITY. CDB: 25 0 0 0 0 0 0 0 0 > > > HARDWARE FAILURE asc:9,ee > > > Vendor Specific ASCQ field > > > replacable unit: 80 sks: 80,0 > > > > > > Disk da0 is my boot disk where FreeBSD is installed and works fine. > > > Disk da1 is only used to store files, but didn't have any files on it > > > at this time. > > > Disk da1 is mounted at startup (line in /etc/fstab) in /disk_da1. > > > > > > Now I can't boot FreeBSD because fsck finds errors on da1. I looked > > > inside and no cables are loose or anything. > > > > > > Is my disk really screwed up just like that, or could it be something > > > else? > > > > The disk is really screwed, generally SCSI messages like this tend > > to be accurate (unfortunately in this case) > > > > > > Well, the disk seems really broken. It's not accessable anymore. > > Is it possible to put a new, modern SCSI-disk in the Alpha? Or is the > SCSI-connection changed. All SCSI drives work in the Alpha (cooling might be an issue in the PWS for hot running drives) -- | / o / /_ _ wilko@FreeBSD.org |/|/ / / /( (_) Bulte Arnhem, the Netherlands To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sun Sep 1 9:23: 1 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 84F0437B400; Sun, 1 Sep 2002 09:22:58 -0700 (PDT) Received: from smtp.web.de (smtp02.web.de [217.72.192.151]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9A64443E4A; Sun, 1 Sep 2002 09:22:57 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [80.129.124.127] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17lXUz-0005GM-00; Sun, 01 Sep 2002 18:22:53 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 187423F6; Sun, 1 Sep 2002 18:22:49 +0200 (CEST) Received: from jan-freebsd.lan (jan-freebsd.lan [192.168.0.22]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id BDF6477; Sun, 1 Sep 2002 18:22:41 +0200 (CEST) Subject: Re: disk hardware failure? => new disk From: Jan Lentfer To: Wilko Bulte Cc: Marco Beishuizen , FreeBSD alpha mailing list , FreeBSD questions mailing list In-Reply-To: <20020901164118.A4222@freebie.xs4all.nl> References: <20020901112635.C3683@freebie.xs4all.nl> <20020901164118.A4222@freebie.xs4all.nl> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 01 Sep 2002 18:22:49 +0200 Message-Id: <1030897369.237.16.camel@jan-freebsd.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Am So, 2002-09-01 um 16.41 schrieb Wilko Bulte: > All SCSI drives work in the Alpha (cooling might be an issue in the PWS > for hot running drives) That is definitely right. The ventilation in the PWS case is not really great. One of my disks died a painful (for me) burning death - it got so hot you couldn't touch it without burning your hand, the disk died after 3 days. I put a 2 fan cooling system in front of the disks, now everything works fine. Jan -- Jan Lentfer System Administrator Molecular Cell Biology / AG Holstein, Darmstadt University of Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 4712037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 6:46:38 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C425937B400 for ; Tue, 3 Sep 2002 06:46:36 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id F366D43E42 for ; Tue, 3 Sep 2002 06:46:35 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id JAA26103; Tue, 3 Sep 2002 09:46:35 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g83Dk5s54886; Tue, 3 Sep 2002 09:46:05 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15732.48413.244116.575687@grasshopper.cs.duke.edu> Date: Tue, 3 Sep 2002 09:46:05 -0400 (EDT) To: Jan Lentfer Cc: freebsd-alpha@FreeBSD.ORG Subject: Re: Running Linux/Alpha Mozilla on FreeBSD/Alpha In-Reply-To: <1030782428.237.10.camel@jan-freebsd.lan> References: <1030742132.8914.13.camel@jan-freebsd.lan> <15727.60644.348194.578769@grasshopper.cs.duke.edu> <1030782428.237.10.camel@jan-freebsd.lan> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Jan Lentfer writes: > http://ftp.mozilla.org/pub/mozilla/releases/mozilla1.0/Red_Hat_7x_RPMS/alpha/mozilla-1.0.0-9.alpha.rpm > http://ftp.mozilla.org/pub/mozilla/releases/mozilla1.0/Red_Hat_7x_RPMS/alpha/mozilla-nspr-1.0.0-9.alpha.rpm > > and moved everything from /usr to /usr/local and edited the Does anybody know if there are mozilla 1.1 linux/alpha binaries available anywhere? Mozilla.org only has 1.0 binaries for linux/alpha. I think the 1.0 version of mozilla is vulnerable to the "huge fonts on a web page can crash your X server" bug, so 1.1 would certainly be better.. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 8:34:12 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9990537B400 for ; Tue, 3 Sep 2002 08:34:10 -0700 (PDT) Received: from sophia.inria.fr (sophia.inria.fr [138.96.64.20]) by mx1.FreeBSD.org (Postfix) with ESMTP id B640D43E6A for ; Tue, 3 Sep 2002 08:34:09 -0700 (PDT) (envelope-from Ludovic.Maitre@sophia.inria.fr) Received: from sophia.inria.fr (ubik.inria.fr [138.96.66.21]) by sophia.inria.fr (8.12.5/8.12.5) with ESMTP id g83FY8Rx031238 for ; Tue, 3 Sep 2002 17:34:08 +0200 X-Authentication-Warning: sophia.inria.fr: Host ubik.inria.fr [138.96.66.21] claimed to be sophia.inria.fr Message-ID: <3D74D670.3FE21B1D@sophia.inria.fr> Date: Tue, 03 Sep 2002 17:34:08 +0200 From: Ludovic Maitre Organization: INRIA Sophia Antipolis X-Mailer: Mozilla 4.76 [en] (X11; U; SunOS 5.6 sun4u) X-Accept-Language: fr, fr-FR, en MIME-Version: 1.0 To: freebsd-alpha@FreeBSD.org Subject: subscribe Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org I want to subscribe (and to obtain FreeBSD 4.6.2 for my alphastation 500) -- Ludovic.Maitre@sophia.inria.fr INRIA - 2004 route des lucioles - BP 93 Tel: (33/0) 4 92 38 50 41 06902 SOPHIA-ANTIPOLIS cedex (France) Fax: (33/0) 4 92 38 76 02 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 9: 5:55 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DD04B37B400 for ; Tue, 3 Sep 2002 09:05:53 -0700 (PDT) Received: from gull.mail.pas.earthlink.net (gull.mail.pas.earthlink.net [207.217.120.84]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8B9C943E75 for ; Tue, 3 Sep 2002 09:05:53 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0292.cvx40-bradley.dialup.earthlink.net ([216.244.43.37] helo=mindspring.com) by gull.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17mGBW-0003pr-00; Tue, 03 Sep 2002 09:05:46 -0700 Message-ID: <3D74DD9C.81E883DE@mindspring.com> Date: Tue, 03 Sep 2002 09:04:44 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Ludovic Maitre Cc: freebsd-alpha@FreeBSD.org Subject: Re: subscribe References: <3D74D670.3FE21B1D@sophia.inria.fr> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Ludovic Maitre wrote: > > I want to subscribe (and to obtain FreeBSD 4.6.2 for my alphastation > 500) send a message to: majordomo@freebsd.org with a body of: help Alpha ISO's are available from: ftp://ftp.fr.FreeBSD.org/pub/FreeBSD/ISO-IMAGES-Alpha/4.6/ Apparently whoever put the text file in the 4.6.2 directory was lying. 8-(. But the 4.6 version is there. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 10: 2:51 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1952037B400 for ; Tue, 3 Sep 2002 10:02:49 -0700 (PDT) Received: from gnah.bolet.org (gnah.bolet.org [80.65.226.87]) by mx1.FreeBSD.org (Postfix) with ESMTP id AFB9F43E6A for ; Tue, 3 Sep 2002 10:02:47 -0700 (PDT) (envelope-from pornin@bolet.org) Received: from gnah.bolet.org (localhost [127.0.0.1]) by gnah.bolet.org (8.12.5/8.12.5) with ESMTP id g83H2ikU048060 for ; Tue, 3 Sep 2002 19:02:45 +0200 (CEST) (envelope-from pornin@bolet.org) Received: (from pornin@localhost) by gnah.bolet.org (8.12.5/8.12.5/Submit) id g83H2giw048059 for alpha@freebsd.org; Tue, 3 Sep 2002 19:02:42 +0200 (CEST) Date: Tue, 3 Sep 2002 19:02:40 +0200 From: Thomas Pornin To: alpha@freebsd.org Subject: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020903190239.A48033@gnah.bolet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Hello, I have at home an AXPpci33 machine running happily FreeBSD-4.6. This is a standard NoName board, with some PCI slots, onboard SCSI, a 166 MHz cpu (21066 I believe) and 32 MBytes of ram. The board features a serial port, which seems to be recognized by FreeBSD: sio0: configured irq 4 not in bitmap of probed irqs 0 sio0 at port 0x3f8-0x3ff irq 4 on isa0 sio0: type 16550A sio0: interrupting at ISA irq 4 sio1: reserved for low-level i/o I shall soon acquire an old IPC Sun SparcStation. Those stations can use a serial console, and are actually supposed to switch to a serial console in the absence of any keyboard (there might be a switch on the board but that's not a problem). If I remember correctly, the DB25 connector is female on those stations, and it is possible that the pin layout is non-standard. What I would like to do is connecting the serial ports of the Alpha and the Sparc machines so that I could control the Sparc console from a shell on the Alpha machine. I have two questions: ** What software should I use on the Alpha box ? I know there exists some program called "minicom", which has a FreeBSD port, that might be adequate. But is this the standard solution ? ** What is the layout of the serial cable to use ? I suspect that there are non-standard serial ports on some Alpha machines for the following reason: I once had a Multia, which is basically a NoName board in a small box. That Multia could use a vt220 as console when no keyboard was connected. A simple null-modem cable worked (those cables sold as "doom-cable", with a DB9 and a DB25 at each end). I tried the same setting on my current Alpha machine, and it did not work. Some sort of reference on serial cables would be much appreciated. There is some info on: http://www.netbsd.org/Documentation/Hardware/Misc/serial.html but it is only partial. Thanks for any answer, --Thomas Pornin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 11:18: 0 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 860D537B405 for ; Tue, 3 Sep 2002 11:17:54 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5FB7343E6E for ; Tue, 3 Sep 2002 11:17:53 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g83IHoJR014229; Tue, 3 Sep 2002 20:17:50 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g83IHnWn014228; Tue, 3 Sep 2002 20:17:49 +0200 (CEST) Date: Tue, 3 Sep 2002 20:17:49 +0200 From: Wilko Bulte To: Thomas Pornin Cc: alpha@FreeBSD.ORG Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020903201749.F14118@freebie.xs4all.nl> References: <20020903190239.A48033@gnah.bolet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20020903190239.A48033@gnah.bolet.org>; from pornin@bolet.org on Tue, Sep 03, 2002 at 07:02:40PM +0200 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 07:02:40PM +0200, Thomas Pornin wrote: > I shall soon acquire an old IPC Sun SparcStation. Those stations can > use a serial console, and are actually supposed to switch to a serial > console in the absence of any keyboard (there might be a switch on the Yep, just omit the keyboard. Works fine (on my SS5, and used to work on my IPC as well). > board but that's not a problem). If I remember correctly, the DB25 > connector is female on those stations, and it is possible that the pin > layout is non-standard. Yep female, but IIRC the pin layout is standard. > What I would like to do is connecting the serial ports of the Alpha > and the Sparc machines so that I could control the Sparc console from > a shell on the Alpha machine. I have two questions: > > ** What software should I use on the Alpha box ? I know there exists > some program called "minicom", which has a FreeBSD port, that might > be adequate. But is this the standard solution ? Whatever you prefer. I tend to use 'tip' which is part of the base system. > ** What is the layout of the serial cable to use ? I suspect that there > are non-standard serial ports on some Alpha machines for the following > reason: I once had a Multia, which is basically a NoName board in a > small box. That Multia could use a vt220 as console when no keyboard > was connected. A simple null-modem cable worked (those cables sold as > "doom-cable", with a DB9 and a DB25 at each end). I tried the same > setting on my current Alpha machine, and it did not work. Hmm, nothing special I remember. The connector on the NoName mainboard has a 'special' pinout compared to what PC mainboards have. -- | / o / /_ _ wilko@FreeBSD.org |/|/ / / /( (_) Bulte Arnhem, the Netherlands To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 11:46:17 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DD7DB37B400 for ; Tue, 3 Sep 2002 11:46:14 -0700 (PDT) Received: from smtp.web.de (smtp03.web.de [217.72.192.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0648643E42 for ; Tue, 3 Sep 2002 11:46:14 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [217.82.44.145] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17mIgj-000557-00; Tue, 03 Sep 2002 20:46:09 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 0997C383; Tue, 3 Sep 2002 20:46:03 +0200 (CEST) Received: from jan-freebsd.lan (jan-freebsd.lan [192.168.0.22]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id A905F1D1; Tue, 3 Sep 2002 20:45:39 +0200 (CEST) Subject: Old machines, not Alpha was:AXPpci33 <-> Sparc sun4c serial cable From: Jan Lentfer To: Wilko Bulte Cc: freebsd-alpha@freebsd.org In-Reply-To: <20020903201749.F14118@freebie.xs4all.nl> References: <20020903190239.A48033@gnah.bolet.org> <20020903201749.F14118@freebie.xs4all.nl> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 03 Sep 2002 20:45:47 +0200 Message-Id: <1031078749.7991.15.camel@jan-freebsd.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Hi all, since the topic already started I'd liked to jump in. Do people on this list also run other vintage (non-alpha, NON-X86!!) machines (not the alpha is all vintage, but mine is). If so which? What experiences? What OSes? I just ask because I had the offer to buy a MicroVax II for EUR 150.- but didn't take it because I was unsure if it was really worth it. Of course I know the hardware isn't worth very much anymore, but what about the fun factor of such machines? I also saw a PDP-11/70 (?? one of those in a almost standard tower case) on American ebay with full docu and licenses. Any one got one of these really old machines as a toy? I know it's off topic again, but.... (can't think of any excuses anymore) Thanks in advance, Jan -- Jan Lentfer System Administrator Molecular Cell Biology / AG Holstein, Darmstadt University of Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 4712037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 12: 0:30 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AC4C337B400 for ; Tue, 3 Sep 2002 12:00:23 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9CC5C43E4A for ; Tue, 3 Sep 2002 12:00:22 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g83J0LJR014628; Tue, 3 Sep 2002 21:00:21 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g83J0Lp1014627; Tue, 3 Sep 2002 21:00:21 +0200 (CEST) Date: Tue, 3 Sep 2002 21:00:21 +0200 From: Wilko Bulte To: Jan Lentfer Cc: freebsd-alpha@freebsd.org Subject: Re: Old machines, not Alpha was:AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020903210021.A14603@freebie.xs4all.nl> References: <20020903190239.A48033@gnah.bolet.org> <20020903201749.F14118@freebie.xs4all.nl> <1031078749.7991.15.camel@jan-freebsd.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <1031078749.7991.15.camel@jan-freebsd.lan>; from Jan.Lentfer@web.de on Tue, Sep 03, 2002 at 08:45:47PM +0200 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 08:45:47PM +0200, Jan Lentfer wrote: Well.. 2.11BSD on my PDP and another BSD rev on my MvaxII do make fun. You better Google for 'PUPS' or "TUHS" for appropriate mailing lists Wilko > Hi all, > > since the topic already started I'd liked to jump in. Do people on this > list also run other vintage (non-alpha, NON-X86!!) machines (not the > alpha is all vintage, but mine is). If so which? What experiences? What > OSes? I just ask because I had the offer to buy a MicroVax II for EUR > 150.- but didn't take it because I was unsure if it was really worth it. > Of course I know the hardware isn't worth very much anymore, but what > about the fun factor of such machines? I also saw a PDP-11/70 (?? one of > those in a almost standard tower case) on American ebay with full docu > and licenses. Any one got one of these really old machines as a toy? > > I know it's off topic again, but.... (can't think of any excuses > anymore) > > Thanks in advance, > > Jan > -- > Jan Lentfer > System Administrator > Molecular Cell Biology / AG Holstein, Darmstadt University of > Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany > Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 > 4712037 ---end of quoted text--- -- | / o / /_ _ wilko@FreeBSD.org |/|/ / / /( (_) Bulte Arnhem, the Netherlands To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 12:42:20 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1F78D37B400 for ; Tue, 3 Sep 2002 12:42:18 -0700 (PDT) Received: from thea.blinkenlights.nl (thea.blinkenlights.nl [62.58.162.229]) by mx1.FreeBSD.org (Postfix) with ESMTP id 37A3E43E72 for ; Tue, 3 Sep 2002 12:42:12 -0700 (PDT) (envelope-from sten@blinkenlights.nl) Received: by thea.blinkenlights.nl (Postfix, from userid 101) id 3F4C339305; Tue, 3 Sep 2002 21:41:50 +0200 (CEST) Received: from localhost (localhost [127.0.0.1]) by thea.blinkenlights.nl (Postfix) with ESMTP id 2FA1DD59C; Tue, 3 Sep 2002 21:41:50 +0200 (CEST) Date: Tue, 3 Sep 2002 21:41:50 +0200 (CEST) From: Sten To: Thomas Pornin Cc: alpha@freebsd.org Subject: Re: AXPpci33 <-> Sparc sun4c serial cable In-Reply-To: <20020903190239.A48033@gnah.bolet.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, 3 Sep 2002, Thomas Pornin wrote: > Hello, > > What I would like to do is connecting the serial ports of the Alpha > and the Sparc machines so that I could control the Sparc console from > a shell on the Alpha machine. I have two questions: > > ** What software should I use on the Alpha box ? I know there exists > some program called "minicom", which has a FreeBSD port, that might > be adequate. But is this the standard solution ? Just use the terminal prog you feel most comfortable with, minicom, kermit for example. > ** What is the layout of the serial cable to use ? I suspect that there > are non-standard serial ports on some Alpha machines for the following > reason: I once had a Multia, which is basically a NoName board in a > small box. That Multia could use a vt220 as console when no keyboard > was connected. A simple null-modem cable worked (those cables sold as > "doom-cable", with a DB9 and a DB25 at each end). I tried the same > setting on my current Alpha machine, and it did not work. The serial ports have standard pin layouts. The only issue with suns is the fact the db25 is female. Just make sure to use the correct speed settings ( normally 9600 8n1 without hardware accel ). -- Sten Spans "What does one do with ones money, when there is no more empty rackspace ?" To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 14:30: 6 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6E8BD37B400 for ; Tue, 3 Sep 2002 14:30:00 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id DEA8F43E65 for ; Tue, 3 Sep 2002 14:29:58 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g83LTs6i049030 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Tue, 3 Sep 2002 23:29:55 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g83LTKFJ082598 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Tue, 3 Sep 2002 23:29:21 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g83LS5mC082579; Tue, 3 Sep 2002 23:28:05 +0200 (CEST)?g (envelope-from ticso) Date: Tue, 3 Sep 2002 23:28:05 +0200 From: Bernd Walter To: Jan Lentfer Cc: Wilko Bulte , freebsd-alpha@FreeBSD.ORG Subject: Re: Old machines, not Alpha was:AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020903212805.GB82356@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <20020903190239.A48033@gnah.bolet.org> <20020903201749.F14118@freebie.xs4all.nl> <1031078749.7991.15.camel@jan-freebsd.lan> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1031078749.7991.15.camel@jan-freebsd.lan> X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 08:45:47PM +0200, Jan Lentfer wrote: > Hi all, > > since the topic already started I'd liked to jump in. Do people on this > list also run other vintage (non-alpha, NON-X86!!) machines (not the > alpha is all vintage, but mine is). If so which? What experiences? What > OSes? I just ask because I had the offer to buy a MicroVax II for EUR > 150.- but didn't take it because I was unsure if it was really worth it. > Of course I know the hardware isn't worth very much anymore, but what > about the fun factor of such machines? I also saw a PDP-11/70 (?? one of > those in a almost standard tower case) on American ebay with full docu > and licenses. Any one got one of these really old machines as a toy? > > I know it's off topic again, but.... (can't think of any excuses > anymore) I'm running a bunch of Sun3 (2x 3/50 2x 3/60 3/110) with NetBSD, a May Basic Four GPx7050 (an SMP 80386 with VME) with their own Unix (Sequent AFAIK), and some sun4m machines under NetBSD and Solaris. Photos on http://www.cosmo-project.de/~bernd/cicely-pics -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 14:52:56 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 735EB37B400 for ; Tue, 3 Sep 2002 14:52:44 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id A6E5943E42 for ; Tue, 3 Sep 2002 14:52:43 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g83Lqb6i049372 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Tue, 3 Sep 2002 23:52:39 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g83LqbFJ082652 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Tue, 3 Sep 2002 23:52:37 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g83LqYsh082651; Tue, 3 Sep 2002 23:52:34 +0200 (CEST)?g (envelope-from ticso) Date: Tue, 3 Sep 2002 23:52:32 +0200 From: Bernd Walter To: Thomas Pornin Cc: alpha@FreeBSD.ORG Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020903215232.GC82356@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <20020903190239.A48033@gnah.bolet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20020903190239.A48033@gnah.bolet.org> X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 07:02:40PM +0200, Thomas Pornin wrote: > Hello, > > I have at home an AXPpci33 machine running happily FreeBSD-4.6. This is > a standard NoName board, with some PCI slots, onboard SCSI, a 166 MHz > cpu (21066 I believe) and 32 MBytes of ram. The board features a serial > port, which seems to be recognized by FreeBSD: > > sio0: configured irq 4 not in bitmap of probed irqs 0 > sio0 at port 0x3f8-0x3ff irq 4 on isa0 > sio0: type 16550A > sio0: interrupting at ISA irq 4 > sio1: reserved for low-level i/o If sio0 is your console and you want to use sio1 you have to remove the debug flags on sio1. > I shall soon acquire an old IPC Sun SparcStation. Those stations can > use a serial console, and are actually supposed to switch to a serial > console in the absence of any keyboard (there might be a switch on the > board but that's not a problem). If I remember correctly, the DB25 > connector is female on those stations, and it is possible that the pin > layout is non-standard. Take a gender changer and you get the same pinout as on a PC DB-25. > What I would like to do is connecting the serial ports of the Alpha > and the Sparc machines so that I could control the Sparc console from > a shell on the Alpha machine. I have two questions: > > ** What software should I use on the Alpha box ? I know there exists > some program called "minicom", which has a FreeBSD port, that might > be adequate. But is this the standard solution ? I prefer tip, but minicom will also do. > ** What is the layout of the serial cable to use ? I suspect that there > are non-standard serial ports on some Alpha machines for the following For console on FreeBSD and Suns a 3 wire null-modem cable will do. For the general use you want a fully connected null-modem which also handles DCD and DSR which many cheap cables don't do. I don't remember the pinout on the NoName board - I can find out if you need to know. But I can say for shure that pinout is different to the pinout used in most PCs! > reason: I once had a Multia, which is basically a NoName board in a Multias are different in many aspects - they only share the same CPU family and therefor the same embedded chipset which make them very similar from the software standpoint. > small box. That Multia could use a vt220 as console when no keyboard The SRM console isn't requiring a special terminal type - I once used an old ADM3A (a real Lear Siegler from 1980). > was connected. A simple null-modem cable worked (those cables sold as > "doom-cable", with a DB9 and a DB25 at each end). I tried the same > setting on my current Alpha machine, and it did not work. What do you mean with did not work? > Some sort of reference on serial cables would be much appreciated. > There is some info on: > http://www.netbsd.org/Documentation/Hardware/Misc/serial.html > but it is only partial. I usually build my cables myself just because I don't have the time to search for the right one. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 16:30:52 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8925B37B400 for ; Tue, 3 Sep 2002 16:30:50 -0700 (PDT) Received: from mail.inka.de (quechua.inka.de [212.227.14.2]) by mx1.FreeBSD.org (Postfix) with ESMTP id 055A843E75 for ; Tue, 3 Sep 2002 16:30:49 -0700 (PDT) (envelope-from mailnull@mips.inka.de) Received: from kemoauc.mips.inka.de (uucp@) by mail.inka.de with local-bsmtp id 17mN8B-0004Oi-01; Wed, 4 Sep 2002 01:30:47 +0200 Received: from kemoauc.mips.inka.de (localhost [127.0.0.1]) by kemoauc.mips.inka.de (8.12.6/8.12.5) with ESMTP id g83MxFcG035125 for ; Wed, 4 Sep 2002 00:59:15 +0200 (CEST) (envelope-from mailnull@localhost.mips.inka.de) Received: (from mailnull@localhost) by kemoauc.mips.inka.de (8.12.6/8.12.6/Submit) id g83MxFDI035124 for freebsd-alpha@freebsd.org; Wed, 4 Sep 2002 00:59:15 +0200 (CEST) From: naddy@mips.inka.de (Christian Weisgerber) Subject: Re: Old machines, not Alpha was:AXPpci33 <-> Sparc sun4c serial cable Date: Tue, 3 Sep 2002 22:59:14 +0000 (UTC) Message-ID: References: <20020903190239.A48033@gnah.bolet.org> <20020903201749.F14118@freebie.xs4all.nl> <1031078749.7991.15.camel@jan-freebsd.lan> Originator: naddy@mips.inka.de (Christian Weisgerber) To: freebsd-alpha@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Jan Lentfer wrote: > since the topic already started I'd liked to jump in. Do people on this > list also run other vintage (non-alpha, NON-X86!!) machines (not the > alpha is all vintage, but mine is). If so which? What experiences? What > OSes? I have a Sun Blade 100 (UltraSPARC) on loan, running OpenBSD. I also used to have a Sun SPARCstation 20/71 but I got tired of that box and after building OpenBSD/sparc packages for three releases in a row I decided I had had enough of that platform. I very much like non-x86 hardware, but I'm not a collector. For me a machine must still be usable. VAXen, m68k machines, non-Ultra SPARCs, alphas < 21164, and other boxes in that speed range don't qualify. YMMV, I know people who very much enjoy what I call "museum hardware", although I have no idea what they actually *do* with it. > I know it's off topic again, but.... (can't think of any excuses > anymore) Maybe continue on -chat? -- Christian "naddy" Weisgerber naddy@mips.inka.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 16:30:57 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 90A9D37B401 for ; Tue, 3 Sep 2002 16:30:50 -0700 (PDT) Received: from mail.inka.de (quechua.inka.de [212.227.14.2]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0541743E65 for ; Tue, 3 Sep 2002 16:30:49 -0700 (PDT) (envelope-from mailnull@mips.inka.de) Received: from kemoauc.mips.inka.de (uucp@) by mail.inka.de with local-bsmtp id 17mN8B-0004Oi-00; Wed, 4 Sep 2002 01:30:47 +0200 Received: from kemoauc.mips.inka.de (localhost [127.0.0.1]) by kemoauc.mips.inka.de (8.12.6/8.12.5) with ESMTP id g83MXgcG034251 for ; Wed, 4 Sep 2002 00:33:42 +0200 (CEST) (envelope-from mailnull@localhost.mips.inka.de) Received: (from mailnull@localhost) by kemoauc.mips.inka.de (8.12.6/8.12.6/Submit) id g83MXg5V034250 for freebsd-alpha@freebsd.org; Wed, 4 Sep 2002 00:33:42 +0200 (CEST) From: naddy@mips.inka.de (Christian Weisgerber) Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Date: Tue, 3 Sep 2002 22:33:41 +0000 (UTC) Message-ID: References: <20020903190239.A48033@gnah.bolet.org> Originator: naddy@mips.inka.de (Christian Weisgerber) To: freebsd-alpha@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Thomas Pornin wrote: > I shall soon acquire an old IPC Sun SparcStation. Those stations can > use a serial console, and are actually supposed to switch to a serial > console in the absence of any keyboard (there might be a switch on the > board but that's not a problem). If I remember correctly, the DB25 > connector is female on those stations, and it is possible that the pin > layout is non-standard. Actually, the IPC uses 8-pin Mini-DIN connectors for its serial ports. http://www.sunhelp.org/faq/sunref2.html > ** What software should I use on the Alpha box ? I know there exists > some program called "minicom", which has a FreeBSD port, that might > be adequate. But is this the standard solution ? Minicom is pretty silly because it combines terminal emulation and communication functions in the style of corresponding PC programs. (It's roughly a clone of Telix for MS-DOS.) On a Unix box you already have a terminal, no need to insert an emulation layer. A simple communication program such as cu(1) or tip(1), or kermit(1) from the ports, will do. > I suspect that there are non-standard serial ports on some Alpha > machines Not that I'm aware of. > for the following reason: I once had a Multia, which is basically > a NoName board in a small box. That Multia could use a vt220 as > console when no keyboard was connected. A simple null-modem cable > worked (those cables sold as "doom-cable", with a DB9 and a DB25 > at each end). Yup, that sounds right. > I tried the same setting on my current Alpha machine, and it did > not work. Which doesn't pin-point the error. -- Christian "naddy" Weisgerber naddy@mips.inka.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 19:50:28 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C818B37B400 for ; Tue, 3 Sep 2002 19:50:26 -0700 (PDT) Received: from dragon.nuxi.com (trang.nuxi.com [66.92.13.169]) by mx1.FreeBSD.org (Postfix) with ESMTP id 505CC43E6A for ; Tue, 3 Sep 2002 19:50:26 -0700 (PDT) (envelope-from obrien@NUXI.com) Received: from dragon.nuxi.com (obrien@localhost [127.0.0.1]) by dragon.nuxi.com (8.12.6/8.12.2) with ESMTP id g842oPfO090404; Tue, 3 Sep 2002 19:50:25 -0700 (PDT) (envelope-from obrien@dragon.nuxi.com) Received: (from obrien@localhost) by dragon.nuxi.com (8.12.6/8.12.5/Submit) id g842oPKp090403; Tue, 3 Sep 2002 19:50:25 -0700 (PDT) Date: Tue, 3 Sep 2002 19:50:25 -0700 From: "David O'Brien" To: Christian Weisgerber Cc: freebsd-alpha@freebsd.org Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020904025025.GA90319@dragon.nuxi.com> Reply-To: obrien@freebsd.org References: <20020903190239.A48033@gnah.bolet.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4i X-Operating-System: FreeBSD 5.0-CURRENT Organization: The NUXI BSD Group X-Pgp-Rsa-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Rsa-Keyid: 1024/34F9F9D5 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 10:33:41PM +0000, Christian Weisgerber wrote: > > ** What software should I use on the Alpha box ? I know there exists > > some program called "minicom", which has a FreeBSD port, that might > > be adequate. But is this the standard solution ? > > Minicom is pretty silly because it combines terminal emulation and > communication functions in the style of corresponding PC programs. > (It's roughly a clone of Telix for MS-DOS.) On a Unix box you > already have a terminal, no need to insert an emulation layer. A > simple communication program such as cu(1) or tip(1), or kermit(1) > from the ports, will do. Surprised no one mentioned conserver -- it is great in that you can can have one person using it, and multiple people watching what is going on. It is nice to be on the console at home, and when I go to work temperary "bump" the home process. But when I return home, I still have the full console output. One can do this with screen(1), but it takes planning that using conserver does not. ALSO conserver logs all console output to a file. So if you wondered what was printed out then the box paniced -- you know. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Tue Sep 3 23:33:39 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 09A1E37B400; Tue, 3 Sep 2002 23:33:37 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id EDCCD43E65; Tue, 3 Sep 2002 23:33:35 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g846XYJR017090; Wed, 4 Sep 2002 08:33:35 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g846XYYG017089; Wed, 4 Sep 2002 08:33:34 +0200 (CEST) Date: Wed, 4 Sep 2002 08:33:34 +0200 From: Wilko Bulte To: "David O'Brien" Cc: Christian Weisgerber , freebsd-alpha@FreeBSD.ORG Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020904083334.E16965@freebie.xs4all.nl> References: <20020903190239.A48033@gnah.bolet.org> <20020904025025.GA90319@dragon.nuxi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20020904025025.GA90319@dragon.nuxi.com>; from obrien@FreeBSD.ORG on Tue, Sep 03, 2002 at 07:50:25PM -0700 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 07:50:25PM -0700, David O'Brien wrote: > On Tue, Sep 03, 2002 at 10:33:41PM +0000, Christian Weisgerber wrote: > > > ** What software should I use on the Alpha box ? I know there exists > > > some program called "minicom", which has a FreeBSD port, that might > > > be adequate. But is this the standard solution ? > > > > Minicom is pretty silly because it combines terminal emulation and > > communication functions in the style of corresponding PC programs. > > (It's roughly a clone of Telix for MS-DOS.) On a Unix box you > > already have a terminal, no need to insert an emulation layer. A > > simple communication program such as cu(1) or tip(1), or kermit(1) > > from the ports, will do. > > Surprised no one mentioned conserver -- it is great in that you can can > have one person using it, and multiple people watching what is going on. > It is nice to be on the console at home, and when I go to work temperary > "bump" the home process. But when I return home, I still have the full > console output. One can do this with screen(1), but it takes planning > that using conserver does not. Yup, I can support David's plug for conserver. We use it at work to control a truckload of serial Alpha consoles. In our case controlling DECserver terminal servers. Very neat. -- | / o / /_ _ |/|/ / / /( (_) Bulte wilko@FreeBSD.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 0:23:38 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7264137B400 for ; Wed, 4 Sep 2002 00:23:36 -0700 (PDT) Received: from gnah.bolet.org (gnah.bolet.org [80.65.226.87]) by mx1.FreeBSD.org (Postfix) with ESMTP id 13FC843E42 for ; Wed, 4 Sep 2002 00:23:35 -0700 (PDT) (envelope-from pornin@bolet.org) Received: from gnah.bolet.org (localhost [127.0.0.1]) by gnah.bolet.org (8.12.5/8.12.5) with ESMTP id g847NTkU057067; Wed, 4 Sep 2002 09:23:30 +0200 (CEST) (envelope-from pornin@bolet.org) Received: (from pornin@localhost) by gnah.bolet.org (8.12.5/8.12.5/Submit) id g847NRQ7057066; Wed, 4 Sep 2002 09:23:28 +0200 (CEST) Date: Wed, 4 Sep 2002 09:23:26 +0200 From: Thomas Pornin To: alpha@FreeBSD.ORG Cc: ticso@cicely.de Subject: Re: AXPpci33 <-> Sparc sun4c serial cable Message-ID: <20020904092325.A56835@gnah.bolet.org> References: <20020903190239.A48033@gnah.bolet.org> <20020903215232.GC82356@cicely5.cicely.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0i In-Reply-To: <20020903215232.GC82356@cicely5.cicely.de>; from ticso@cicely5.cicely.de on Tue, Sep 03, 2002 at 11:52:32PM +0200 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Tue, Sep 03, 2002 at 11:52:32PM +0200, Bernd Walter wrote: > > was connected. A simple null-modem cable worked (those cables sold as > > "doom-cable", with a DB9 and a DB25 at each end). I tried the same > > setting on my current Alpha machine, and it did not work. > > What do you mean with did not work? No I/O on the VT220: no output on the screen, and nothing I typed went to the machine. I used a VGA card (S3 Trio64V+) and a standard PS/2 keyboard, and this worked. I tried several combinations and the behaviour (with or without keyboard) was much similar to what I could get with the Multia, except that the VT remained blank. I tried a "set console serial", to no avail. I cannot do much testing now, since the box is my firewall/primary DNS/mail server, and I gave the VT220 to somebody else. Anyway, thanks for all the answers. When I get the IPC (hopefully in two or three weeks) I will try this out. --Thomas Pornin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 6: 5: 4 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 9E49537B401 for ; Wed, 4 Sep 2002 06:05:00 -0700 (PDT) Received: from mail.speakeasy.net (mail14.speakeasy.net [216.254.0.214]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2ED6D43E75 for ; Wed, 4 Sep 2002 06:04:58 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 2377 invoked from network); 4 Sep 2002 13:05:18 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail14.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 4 Sep 2002 13:05:18 -0000 Received: from laptop.baldwin.cx (laptop.baldwin.cx [192.168.0.4]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g84D4tBv007516; Wed, 4 Sep 2002 09:04:55 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 Date: Wed, 04 Sep 2002 09:04:55 -0400 (EDT) From: John Baldwin To: alpha@FreeBSD.org, peter@FreeBSD.org Subject: FYI: Buffer cache fix gave major speedup Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On the DS20 I have here, Peter's fix to basically enable the buffer cache gave me about a 38% performance increase for a buildworld -j 2 on current: Before: -------------------------------------- build started at 15:11:08 on 08/28/02 build finished at 18:43:05 on 08/28/02 -------------------------------------- Which is a total time of 3:31:57 After: -------------------------------------- build started at 22:41:31 on 09/03/02 build finished at 00:51:46 on 09/04/02 -------------------------------------- Which is a total time of 2:10:15 If this bug exists in 4.x we should really fix it there as well. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 6:17:25 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 422B137B400; Wed, 4 Sep 2002 06:17:24 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1392043E65; Wed, 4 Sep 2002 06:17:23 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g84DHLJR017752; Wed, 4 Sep 2002 15:17:21 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g84DHKEC017751; Wed, 4 Sep 2002 15:17:20 +0200 (CEST) Date: Wed, 4 Sep 2002 15:17:20 +0200 From: Wilko Bulte To: John Baldwin Cc: alpha@FreeBSD.ORG, peter@FreeBSD.ORG Subject: Re: FYI: Buffer cache fix gave major speedup Message-ID: <20020904151720.A17724@freebie.xs4all.nl> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from jhb@FreeBSD.ORG on Wed, Sep 04, 2002 at 09:04:55AM -0400 X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 04, 2002 at 09:04:55AM -0400, John Baldwin wrote: > On the DS20 I have here, Peter's fix to basically enable the buffer cache > gave me about a 38% performance increase for a buildworld -j 2 on current: WOW! That makes Alpha quite a bit faster than before. Buildworlds always felt slow on Alpha to me. -- | / o / /_ _ wilko@FreeBSD.org |/|/ / / /( (_) Bulte Arnhem, the Netherlands To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 6:43:23 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F077B37B400; Wed, 4 Sep 2002 06:43:19 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id D0E0443E75; Wed, 4 Sep 2002 06:43:15 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id JAA10322; Wed, 4 Sep 2002 09:43:15 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g84DgiY56510; Wed, 4 Sep 2002 09:42:44 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.3540.941843.110159@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 09:42:44 -0400 (EDT) To: John Baldwin Cc: alpha@FreeBSD.ORG, peter@FreeBSD.ORG Subject: Re: FYI: Buffer cache fix gave major speedup In-Reply-To: References: X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > On the DS20 I have here, Peter's fix to basically enable the buffer cache > gave me about a 38% performance increase for a buildworld -j 2 on current: > > Before: > -------------------------------------- > build started at 15:11:08 on 08/28/02 > build finished at 18:43:05 on 08/28/02 > -------------------------------------- > Which is a total time of 3:31:57 > > After: > -------------------------------------- > build started at 22:41:31 on 09/03/02 > build finished at 00:51:46 on 09/04/02 > -------------------------------------- > Which is a total time of 2:10:15 > > If this bug exists in 4.x we should really fix it there as well. > Are you talking about btoc()? I think stable is immune. A simple test shows the buffer cache is working just fine there. The problem was that phk's version of btoc was truncating things by using a 32-bit cast. Eg, phk replaced this: #define btoc(x) (((x) + PAGE_MASK) >> PAGE_SHIFT) With this: #define btoc(x) (((unsigned)(x)+PAGE_MASK)>>PAGE_SHIFT) ^^^^^^^^^ Which Peter replaced with: #define btoc(x) (((vm_offset_t)(x)+PAGE_MASK)>>PAGE_SHIFT) So the difference between -stable and -current is now just the (vm_offset_t) cast. I don't _think_ that should make a difference. But I'm afraid I don't understand how the cast caused such an innocuous problem as disabling the buffer cache. I'd have thought it would have caused memory corruption. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 6:46:46 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 654FA37B400; Wed, 4 Sep 2002 06:46:44 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 86D8443E75; Wed, 4 Sep 2002 06:46:43 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id JAA10368; Wed, 4 Sep 2002 09:46:43 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g84DkCb56518; Wed, 4 Sep 2002 09:46:12 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.3748.114705.311952@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 09:46:12 -0400 (EDT) To: John Baldwin Cc: alpha@FreeBSD.ORG Subject: Re: FYI: Buffer cache fix gave major speedup In-Reply-To: References: X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org BTW, I've been following Matt's rlim enforcement saga with only half an ear. Does a stock -current kernel work on alpha these days? Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 7: 1:20 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 03FA437B400 for ; Wed, 4 Sep 2002 07:01:17 -0700 (PDT) Received: from mail.speakeasy.net (mail13.speakeasy.net [216.254.0.213]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9522943E65 for ; Wed, 4 Sep 2002 07:01:16 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 2138 invoked from network); 4 Sep 2002 14:01:15 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail13.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 4 Sep 2002 14:01:15 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g84E1EBv007664; Wed, 4 Sep 2002 10:01:14 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15734.3540.941843.110159@grasshopper.cs.duke.edu> Date: Wed, 04 Sep 2002 10:01:14 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: FYI: Buffer cache fix gave major speedup Cc: peter@FreeBSD.ORG, alpha@FreeBSD.ORG Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 04-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > On the DS20 I have here, Peter's fix to basically enable the buffer cache > > gave me about a 38% performance increase for a buildworld -j 2 on current: > > > > Before: > > -------------------------------------- > > build started at 15:11:08 on 08/28/02 > > build finished at 18:43:05 on 08/28/02 > > -------------------------------------- > > Which is a total time of 3:31:57 > > > > After: > > -------------------------------------- > > build started at 22:41:31 on 09/03/02 > > build finished at 00:51:46 on 09/04/02 > > -------------------------------------- > > Which is a total time of 2:10:15 > > > > If this bug exists in 4.x we should really fix it there as well. > > > > Are you talking about btoc()? I think stable is immune. A simple > test shows the buffer cache is working just fine there. Ah, right. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 7: 1:22 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6232B37B401 for ; Wed, 4 Sep 2002 07:01:19 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id 05F2043E65 for ; Wed, 4 Sep 2002 07:01:19 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 22438 invoked from network); 4 Sep 2002 14:01:16 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 4 Sep 2002 14:01:16 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g84E1GBv007667; Wed, 4 Sep 2002 10:01:16 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15734.3748.114705.311952@grasshopper.cs.duke.edu> Date: Wed, 04 Sep 2002 10:01:16 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: FYI: Buffer cache fix gave major speedup Cc: alpha@FreeBSD.ORG Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 04-Sep-2002 Andrew Gallatin wrote: > > > BTW, I've been following Matt's rlim enforcement saga with only half > an ear. Does a stock -current kernel work on alpha these days? Yes, it should finally work now. -stable kernels should also start working again soon. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 7:11:46 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0C43137B400; Wed, 4 Sep 2002 07:11:44 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id E47D543E65; Wed, 4 Sep 2002 07:11:31 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id KAA11213; Wed, 4 Sep 2002 10:11:30 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g84EB0v56546; Wed, 4 Sep 2002 10:11:00 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.5236.873517.46531@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 10:11:00 -0400 (EDT) To: John Baldwin Cc: peter@FreeBSD.ORG, alpha@FreeBSD.ORG Subject: Re: FYI: Buffer cache fix gave major speedup In-Reply-To: References: <15734.3540.941843.110159@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > > > > > If this bug exists in 4.x we should really fix it there as well. > > > > > > > Are you talking about btoc()? I think stable is immune. A simple > > test shows the buffer cache is working just fine there. > > Ah, right. Maybe not.. The same simple test on a mid-july -current shows the buffer cache works there too, I assume: <10:07am>monet/gallatin:tmp>du -hs zot 200M zot <10:07am>monet/gallatin:tmp>dd if=zot of=/dev/null bs=64k 3200+0 records in 3200+0 records out 209715200 bytes transferred in 1.871934 secs (112031304 bytes/sec) (and iostat shows no disk activity) I guess I'm confused.. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 8:31:48 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7B6BF37B400 for ; Wed, 4 Sep 2002 08:31:45 -0700 (PDT) Received: from mail.speakeasy.net (mail11.speakeasy.net [216.254.0.211]) by mx1.FreeBSD.org (Postfix) with ESMTP id 66D9443E72 for ; Wed, 4 Sep 2002 08:31:44 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 17007 invoked from network); 4 Sep 2002 15:31:44 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail11.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 4 Sep 2002 15:31:44 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g84FVgBv007934; Wed, 4 Sep 2002 11:31:43 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15734.5236.873517.46531@grasshopper.cs.duke.edu> Date: Wed, 04 Sep 2002 11:31:42 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: FYI: Buffer cache fix gave major speedup Cc: alpha@FreeBSD.ORG, peter@FreeBSD.ORG Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 04-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > > > > > > > If this bug exists in 4.x we should really fix it there as well. > > > > > > > > > > Are you talking about btoc()? I think stable is immune. A simple > > > test shows the buffer cache is working just fine there. > > > > Ah, right. > > Maybe not.. The same simple test on a mid-july -current shows the > buffer cache works there too, I assume: > > <10:07am>monet/gallatin:tmp>du -hs zot > 200M zot > <10:07am>monet/gallatin:tmp>dd if=zot of=/dev/null bs=64k > 3200+0 records in > 3200+0 records out > 209715200 bytes transferred in 1.871934 secs (112031304 bytes/sec) > > (and iostat shows no disk activity) > > I guess I'm confused.. Hmm, using top, I have 0 cache during a cvs up. Mem: 17M Active, 10M Inact, 46M Wired, 51M Buf, 929M Free Hmm, maybe that is ok then? stable has vm.iodirenable on so maybe that's why I don't have any 'cache' and a large 'buf' instead. > Drew -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 8:51:16 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C7F5A37B400; Wed, 4 Sep 2002 08:51:12 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 253D343E65; Wed, 4 Sep 2002 08:49:27 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g84FmA6i058912 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Wed, 4 Sep 2002 17:48:12 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g84FmAUV087988 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Wed, 4 Sep 2002 17:48:10 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g84Fm9dR087987; Wed, 4 Sep 2002 17:48:09 +0200 (CEST)?g (envelope-from ticso) Date: Wed, 4 Sep 2002 17:48:09 +0200 From: Bernd Walter To: John Baldwin Cc: Andrew Gallatin , alpha@FreeBSD.org Subject: Re: FYI: Buffer cache fix gave major speedup Message-ID: <20020904154808.GA87724@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <15734.3748.114705.311952@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 04, 2002 at 10:01:16AM -0400, John Baldwin wrote: > > On 04-Sep-2002 Andrew Gallatin wrote: > > > > > > BTW, I've been following Matt's rlim enforcement saga with only half > > an ear. Does a stock -current kernel work on alpha these days? > > Yes, it should finally work now. -stable kernels should also start > working again soon. I'm running -current from today and havn't seen this kind of problem anymore. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 8:51:42 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4169837B400; Wed, 4 Sep 2002 08:51:39 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id CFD2F43E72; Wed, 4 Sep 2002 08:51:15 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g84Fp06i058937 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Wed, 4 Sep 2002 17:51:02 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g84Fp0UV088002 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Wed, 4 Sep 2002 17:51:00 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g84Fp0b7088001; Wed, 4 Sep 2002 17:51:00 +0200 (CEST)?g (envelope-from ticso) Date: Wed, 4 Sep 2002 17:50:59 +0200 From: Bernd Walter To: John Baldwin Cc: Andrew Gallatin , alpha@FreeBSD.org, peter@FreeBSD.org Subject: Re: FYI: Buffer cache fix gave major speedup Message-ID: <20020904155059.GB87724@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <15734.5236.873517.46531@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 04, 2002 at 11:31:42AM -0400, John Baldwin wrote: > > On 04-Sep-2002 Andrew Gallatin wrote: > > > > John Baldwin writes: > > > > > > > > > > If this bug exists in 4.x we should really fix it there as well. > > > > > > > > > > > > > Are you talking about btoc()? I think stable is immune. A simple > > > > test shows the buffer cache is working just fine there. > > > > > > Ah, right. > > > > Maybe not.. The same simple test on a mid-july -current shows the > > buffer cache works there too, I assume: > > > > <10:07am>monet/gallatin:tmp>du -hs zot > > 200M zot > > <10:07am>monet/gallatin:tmp>dd if=zot of=/dev/null bs=64k > > 3200+0 records in > > 3200+0 records out > > 209715200 bytes transferred in 1.871934 secs (112031304 bytes/sec) > > > > (and iostat shows no disk activity) > > > > I guess I'm confused.. > > Hmm, using top, I have 0 cache during a cvs up. > > Mem: 17M Active, 10M Inact, 46M Wired, 51M Buf, 929M Free > > Hmm, maybe that is ok then? stable has vm.iodirenable on so > maybe that's why I don't have any 'cache' and a large 'buf' > instead. This is on cvsup.de.freebsd.org: [93]srv1.cosmo-project.de# uname -a FreeBSD srv1.cosmo-project.de 4.6-STABLE FreeBSD 4.6-STABLE #2: Fri Aug 2 17:38:22 CEST 2002 bernd@srv1.cosmo-project.de:/var/d1/FreeBSD-2002-07-31/src/sys/compile/SRV1 alpha [94]srv1.cosmo-project.de# top | grep Mem Mem: 64M Active, 5688K Inact, 37M Wired, 12M Cache, 22M Buf, 680K Free -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 13:59:52 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DEBBB37B400; Wed, 4 Sep 2002 13:59:44 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5717B43E3B; Wed, 4 Sep 2002 13:59:40 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA28943; Wed, 4 Sep 2002 16:59:39 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g84Kx9N56895; Wed, 4 Sep 2002 16:59:09 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.29725.515274.183629@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 16:59:09 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@freebsd.org Subject: alpha performance on -current In-Reply-To: References: X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > On the DS20 I have here, Peter's fix to basically enable the buffer cache > gave me about a 38% performance increase for a buildworld -j 2 on current: > > Before: > -------------------------------------- > build started at 15:11:08 on 08/28/02 > build finished at 18:43:05 on 08/28/02 > -------------------------------------- > Which is a total time of 3:31:57 > > After: > -------------------------------------- > build started at 22:41:31 on 09/03/02 > build finished at 00:51:46 on 09/04/02 > -------------------------------------- > Which is a total time of 2:10:15 A buildworld from July: 9611.42 real 6149.87 user 2613.20 sys A buildworld today: 8699.25 real 6985.64 user 1379.72 sys For all I know, the speedup is just from disabling WITNESS and INVARIANTS. Speaking of performance, I ran lmbench on my xp1000 under -stable and -current. I built the binary with compaq cc and ran the same binaries on -current and -stable. Both kernels were built without WITNESS, INVARIANTS and DIAGNOSTIC. On -current, I manually created an /etc/malloc.conf symlink to remove malloc debugging. I included results from Tru64 5.1A for comparison. The disk that -current is on is a little faster than the disks -stable and Tru64 are on, that's the only area where its not an apples-to-apples test. The first thing that stands out is that syscalls are *much* more expensive on -current. Nearly a factor of 4 for a null syscall (0.57us -> 2.06 us). I suppose it equates to a latency of ~0.35us for each mutex taken/released. Can this be right? The lmbench null syscall is getppid: # ./lat_syscall null Simple syscall: 2.0178 microseconds # ./lat_syscall null Simple syscall: 2.0333 microseconds # sysctl -w kern.giant.proc=0 kern.giant.proc: 1 -> 0 # ./lat_syscall null Simple syscall: 1.6360 microseconds # ./lat_syscall null Simple syscall: 1.6333 microseconds Is the locking overhead this bad on x86? It looks downright embarrassing on alpha. Can anything be done about it? Are the memory barriers in atomic_cmpset_acq_* really needed? They have the look of belt & suspenders code.. FWIW, The appended diff to remove them reducess null system call latency to 1.6us with kern.giant.proc=1, and 1.4us with kern.giant.proc=0. I'm about to start a buildworld with it, but I don't have any SMP boxes. On the other hand, the pipe results are tremendous. Pipe now goes like a bat out of hell. Congrats to whoever did that. Drew L M B E N C H 2 . 0 S U M M A R Y ------------------------------------ Basic system parameters ---------------------------------------------------- Host OS Description Mhz --------- ------------- ----------------------- ---- monet FreeBSD 4.5-S alpha-freebsd4.5 497 monet FreeBSD 5.0-C alpha-freebsd5.0 497 monet OSF1 V5.1 alphaev6-dec-osf5.1 499 Processor, Processes - times in microseconds - smaller is better ---------------------------------------------------------------- Host OS Mhz null null open selct sig sig fork exec sh call I/O stat clos TCP inst hndl proc proc proc --------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ---- monet FreeBSD 4.5-S 497 0.57 3.78 16.2 22.5 33.7 1.29 6.25 1434 4692 12.K monet FreeBSD 5.0-C 497 2.06 7.52 28.1 44.1 18.7 2.35 8.91 1440 4716 8895 monet OSF1 V5.1 499 0.41 1.33 144. 157. 19.8 1.01 4.56 1039 2745 7611 Context switching - times in microseconds - smaller is better ------------------------------------------------------------- Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw --------- ------------- ----- ------ ------ ------ ------ ------- ------- monet FreeBSD 4.5-S 1.850 20.4 98.1 36.9 139.2 39.7 139.7 monet FreeBSD 5.0-C 4.250 13.6 47.1 22.9 62.2 27.2 68.3 monet OSF1 V5.1 5.430 9.8900 45.2 18.1 49.5 20.2 53.6 *Local* Communication latencies in microseconds - smaller is better ------------------------------------------------------------------- Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP ctxsw UNIX UDP TCP conn --------- ------------- ----- ----- ---- ----- ----- ----- ----- ---- monet FreeBSD 4.5-S 1.850 12.0 14.4 38.5 79.1 43.0 100.0 231. monet FreeBSD 5.0-C 4.250 28.4 33.5 87.5 300. monet OSF1 V5.1 5.430 24.8 48.7 90.3 152.8 84.8 197.4 File & VM system latencies in microseconds - smaller is better -------------------------------------------------------------- Host OS 0K File 10K File Mmap Prot Page Create Delete Create Delete Latency Fault Fault --------- ------------- ------ ------ ------ ------ ------- ----- ----- monet FreeBSD 4.5-S 80.4 57.7 2457.0 1243.8 2980.0 0.527 monet FreeBSD 5.0-C 120.2 99.4 330.3 4854.4 4688.0 0.145 monet OSF1 V5.1 488.3 855.4 1773.0 1083.4 1462.0 2.931 4470.0 *Local* Communication bandwidths in MB/s - bigger is better ----------------------------------------------------------- Host OS Pipe AF TCP File Mmap Bcopy Bcopy Mem Mem UNIX reread reread (libc) (hand) read write --------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- ----- monet FreeBSD 4.5-S 82.3 201. 70.2 159.1 953.5 452.6 342.3 953. 351.2 monet FreeBSD 5.0-C 429. 127. 102. 250.9 946.6 439.3 339.3 947. 350.7 monet OSF1 V5.1 301. 237. 318.1 979.5 451.7 350.1 978. 356.7 Memory latencies in nanoseconds - smaller is better (WARNING - may not be correct, check graphs) --------------------------------------------------- Host OS Mhz L1 $ L2 $ Main mem Guesses --------- ------------- ---- ----- ------ -------- ------- monet FreeBSD 4.5-S 497 6.016 30.1 196.6 monet FreeBSD 5.0-C 497 6.039 30.3 197.7 monet OSF1 V5.1 499 5.859 29.3 194.3 Index: atomic.h =================================================================== RCS file: /home/ncvs/src/sys/alpha/include/atomic.h,v retrieving revision 1.14 diff -u -r1.14 atomic.h --- atomic.h 17 May 2002 05:45:39 -0000 1.14 +++ atomic.h 4 Sep 2002 20:37:43 -0000 @@ -419,14 +419,14 @@ int retval; retval = atomic_cmpset_32(p, cmpval, newval); - alpha_mb(); +/* alpha_mb();*/ return (retval); } static __inline u_int32_t atomic_cmpset_rel_32(volatile u_int32_t *p, u_int32_t cmpval, u_int32_t newval) { - alpha_mb(); +/* alpha_mb();*/ return (atomic_cmpset_32(p, cmpval, newval)); } @@ -436,14 +436,14 @@ int retval; retval = atomic_cmpset_64(p, cmpval, newval); - alpha_mb(); +/* alpha_mb();*/ return (retval); } static __inline u_int64_t atomic_cmpset_rel_64(volatile u_int64_t *p, u_int64_t cmpval, u_int64_t newval) { - alpha_mb(); +/* alpha_mb();*/ return (atomic_cmpset_64(p, cmpval, newval)); } To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 15:33: 8 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3231737B400; Wed, 4 Sep 2002 15:33:06 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1B48743E3B; Wed, 4 Sep 2002 15:33:05 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g84MWv6i062463 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 00:33:01 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g84MWuUV090159 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 00:32:57 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g84MWuxH090158; Thu, 5 Sep 2002 00:32:56 +0200 (CEST)?g (envelope-from ticso) Date: Thu, 5 Sep 2002 00:32:56 +0200 From: Bernd Walter To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current Message-ID: <20020904223255.GI87724@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <15734.29725.515274.183629@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15734.29725.515274.183629@grasshopper.cs.duke.edu> X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 04, 2002 at 04:59:09PM -0400, Andrew Gallatin wrote: > > # ./lat_syscall null > Simple syscall: 2.0178 microseconds > # ./lat_syscall null > Simple syscall: 2.0333 microseconds > # sysctl -w kern.giant.proc=0 > kern.giant.proc: 1 -> 0 > # ./lat_syscall null > Simple syscall: 1.6360 microseconds > # ./lat_syscall null > Simple syscall: 1.6333 microseconds > > > Is the locking overhead this bad on x86? It looks downright > embarrassing on alpha. Can anything be done about it? Are the > memory barriers in atomic_cmpset_acq_* really needed? They have the > look of belt & suspenders code.. They are needed in _acq_ and _rel_ because they are used to build mutexes. You even need them on UP systems to enshure instruction order. Some barriers can be removed in other atomic functions or replaced with write barriers. Currently our alpha_wmb() is mappend to mb, which is done because it's also done in NetBSD - possibly as a workaround for a unnamed problem. > FWIW, The appended diff to remove them reducess null system call > latency to 1.6us with kern.giant.proc=1, and 1.4us with > kern.giant.proc=0. I'm about to start a buildworld with it, but I > don't have any SMP boxes. What kind of CPU is in your XP1000? On >= 21164 the CPU can be initialised into UP operation so that mb and wmb fall back to just an contraint on instruction order. Also locked instructions can be handled inside the CPU in that case. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 16:50:36 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7CB4E37B400; Wed, 4 Sep 2002 16:50:32 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8B77643E6E; Wed, 4 Sep 2002 16:50:31 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id TAA04182; Wed, 4 Sep 2002 19:50:30 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g84No0D57025; Wed, 4 Sep 2002 19:50:00 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.39976.326598.176742@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 19:50:00 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: <20020904223255.GI87724@cicely5.cicely.de> References: <15734.29725.515274.183629@grasshopper.cs.duke.edu> <20020904223255.GI87724@cicely5.cicely.de> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter writes: > On Wed, Sep 04, 2002 at 04:59:09PM -0400, Andrew Gallatin wrote: > > > > # ./lat_syscall null > > Simple syscall: 2.0178 microseconds > > # ./lat_syscall null > > Simple syscall: 2.0333 microseconds > > # sysctl -w kern.giant.proc=0 > > kern.giant.proc: 1 -> 0 > > # ./lat_syscall null > > Simple syscall: 1.6360 microseconds > > # ./lat_syscall null > > Simple syscall: 1.6333 microseconds > > > > > > Is the locking overhead this bad on x86? It looks downright > > embarrassing on alpha. Can anything be done about it? Are the > > memory barriers in atomic_cmpset_acq_* really needed? They have the > > look of belt & suspenders code.. > > They are needed in _acq_ and _rel_ because they are used to build > mutexes. atomic_cmpset_acq_64 calls atomic_cmpset_64() followed by a memory barrier.. atomic_cmpset_64() ends with a memory barrier. So isn't the memory barrier in atomic_cmpset_acq_64() extranious? Eg, you have 2 memory barriers back to back. At any rate, the overhead just plain stinks. At nearly 0.5 us per mutex, they are just way too expensive. > You even need them on UP systems to enshure instruction order. Alphas older than ev6 are in-order processors with regard to instruction order, so presumably they are not needed on ev56 and older machines? > Some barriers can be removed in other atomic functions or replaced > with write barriers. > Currently our alpha_wmb() is mappend to mb, which is done because it's > also done in NetBSD - possibly as a workaround for a unnamed problem. > > > FWIW, The appended diff to remove them reducess null system call > > latency to 1.6us with kern.giant.proc=1, and 1.4us with > > kern.giant.proc=0. I'm about to start a buildworld with it, but I > > don't have any SMP boxes. > > What kind of CPU is in your XP1000? 21264. > On >= 21164 the CPU can be initialised into UP operation so that mb > and wmb fall back to just an contraint on instruction order. > Also locked instructions can be handled inside the CPU in that case. Please elaborate. How does this work with respect to devices? That sounds quite scary for the atomic code in general. FWIW, my machine survived the buildworld. The removal of the memory barriers chopped nearly 7 minutes off the time: with MB: 8699.25 real 6985.64 user 1379.72 sys without MB: 8298.44 real 7010.03 user 969.02 sys Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 18:30:41 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4137037B401; Wed, 4 Sep 2002 18:30:38 -0700 (PDT) Received: from canning.wemm.org (canning.wemm.org [192.203.228.65]) by mx1.FreeBSD.org (Postfix) with ESMTP id F41AF43E6A; Wed, 4 Sep 2002 18:30:37 -0700 (PDT) (envelope-from peter@wemm.org) Received: from wemm.org (localhost [127.0.0.1]) by canning.wemm.org (Postfix) with ESMTP id 1D4B42A88D; Wed, 4 Sep 2002 18:30:34 -0700 (PDT) (envelope-from peter@wemm.org) X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4 To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@freebsd.org Subject: Re: alpha performance on -current In-Reply-To: <15734.29725.515274.183629@grasshopper.cs.duke.edu> Date: Wed, 04 Sep 2002 18:30:34 -0700 From: Peter Wemm Message-Id: <20020905013034.1D4B42A88D@canning.wemm.org> Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Andrew Gallatin wrote: > [non-null null syscall tests] We really need a null system call to test for these. How about __sysnull(2)? What lmbench is testing is syscall overhead *plus* several iterations of mutex gain/releases. Possibly with a couple of sysctl knobs to enable turning on of "typical" locking operations. eg: kern.sysnull.giant = gain/release giant, kern.sysnull.proc = gain/release proc lock etc. Then we can easily test the *actual* syscall overhead. > Is the locking overhead this bad on x86? It looks downright > embarrassing on alpha. Can anything be done about it? Are the > memory barriers in atomic_cmpset_acq_* really needed? They have the > look of belt & suspenders code.. The mutex primitives are one the first things that will get serious optimization attention in order to squeeze every last erg out of the kernel when the time is right. That probably means doing away with the debugging hooks, having a place for MD implementations, etc. There are no shortage of eyes available to do this work, but right now what we have works and we most desperately need to not add another possible source of problems. We still haven't figured out why ithread preemption blows up on the alpha kernel, for example. This really needs to be identified. Given that the mutex stuff has pretty much settled down now, I think we could get away with optional 'fast-as-possible' mutexes now. I know there was some pending preemption and priority propagation work in progress that (I think) affected the implementation and data structures, but I think that's been shelved for now due to ongoing conflicts with KSE. Personally, I'd ship 5.0-R with a kernel option MUTEX_NODEBUG or something like that that strips them down and uses the fast MD implementations if they are available. Naturally this would be mutually exclusive with INVARIANTS, WITNESS etc. Leaving out MUTEX_NODEBUG would revert to the 'safe' version (just without the invariatns assertions etc). I'd probably enable this in GENERIC even. But, this is something for the re's to decide, not me. I know John has pretty strong opinions here so I will not try to speak for him. Cheers, -Peter -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 18:58: 0 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7F6A637B400; Wed, 4 Sep 2002 18:57:58 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9682043E3B; Wed, 4 Sep 2002 18:57:57 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id VAA07341; Wed, 4 Sep 2002 21:57:56 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g851vQW57123; Wed, 4 Sep 2002 21:57:26 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15734.47622.855983.90466@grasshopper.cs.duke.edu> Date: Wed, 4 Sep 2002 21:57:26 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: <15734.39976.326598.176742@grasshopper.cs.duke.edu> References: <15734.29725.515274.183629@grasshopper.cs.duke.edu> <20020904223255.GI87724@cicely5.cicely.de> <15734.39976.326598.176742@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Andrew Gallatin writes: > FWIW, my machine survived the buildworld. The removal of the memory > barriers chopped nearly 7 minutes off the time: As well as an (untimed) -j16 buildworld. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Wed Sep 4 21:50:46 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C71D537B400; Wed, 4 Sep 2002 21:50:43 -0700 (PDT) Received: from hawk.mail.pas.earthlink.net (hawk.mail.pas.earthlink.net [207.217.120.22]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6441143E6E; Wed, 4 Sep 2002 21:50:43 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0312.cvx40-bradley.dialup.earthlink.net ([216.244.43.57] helo=mindspring.com) by hawk.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17mob4-00056M-00; Wed, 04 Sep 2002 21:50:27 -0700 Message-ID: <3D76E256.6A200E16@mindspring.com> Date: Wed, 04 Sep 2002 21:49:26 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Peter Wemm Cc: Andrew Gallatin , John Baldwin , freebsd-alpha@freebsd.org Subject: Re: alpha performance on -current References: <20020905013034.1D4B42A88D@canning.wemm.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Peter Wemm wrote: > Andrew Gallatin wrote: > > [non-null null syscall tests] > > We really need a null system call to test for these. How about > __sysnull(2)? What lmbench is testing is syscall overhead *plus* several > iterations of mutex gain/releases. Possibly with a couple of sysctl knobs > to enable turning on of "typical" locking operations. eg: > kern.sysnull.giant = gain/release giant, kern.sysnull.proc = gain/release > proc lock etc. Then we can easily test the *actual* syscall overhead. Isn't this a reasonable thing for it to test? Or are there system calls that don't have the several iterations of mutex gain/release? IMO, the test should be "all code that is common to all system calls". If this yields bad numbers for now, that's fine, as long as it gets optimized at some point in the future, there should be no problem with the numbers indicating what needs to be optimized. At least that way, the benchmark would have *some* real meaning. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 1:48:47 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1378C37B400; Thu, 5 Sep 2002 01:48:41 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0E22843E4A; Thu, 5 Sep 2002 01:48:40 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g858mR6i068685 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 10:48:31 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g858mQUV092712 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 10:48:26 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g858mPtQ092711; Thu, 5 Sep 2002 10:48:25 +0200 (CEST)?g (envelope-from ticso) Date: Thu, 5 Sep 2002 10:48:25 +0200 From: Bernd Walter To: Andrew Gallatin Cc: ticso@cicely.de, John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current Message-ID: <20020905084824.GJ87724@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <15734.29725.515274.183629@grasshopper.cs.duke.edu> <20020904223255.GI87724@cicely5.cicely.de> <15734.39976.326598.176742@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15734.39976.326598.176742@grasshopper.cs.duke.edu> X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Wed, Sep 04, 2002 at 07:50:00PM -0400, Andrew Gallatin wrote: > > Bernd Walter writes: > > On Wed, Sep 04, 2002 at 04:59:09PM -0400, Andrew Gallatin wrote: > > > > > > # ./lat_syscall null > > > Simple syscall: 2.0178 microseconds > > > # ./lat_syscall null > > > Simple syscall: 2.0333 microseconds > > > # sysctl -w kern.giant.proc=0 > > > kern.giant.proc: 1 -> 0 > > > # ./lat_syscall null > > > Simple syscall: 1.6360 microseconds > > > # ./lat_syscall null > > > Simple syscall: 1.6333 microseconds > > > > > > > > > Is the locking overhead this bad on x86? It looks downright > > > embarrassing on alpha. Can anything be done about it? Are the > > > memory barriers in atomic_cmpset_acq_* really needed? They have the > > > look of belt & suspenders code.. > > > > They are needed in _acq_ and _rel_ because they are used to build > > mutexes. > > atomic_cmpset_acq_64 calls atomic_cmpset_64() followed by a memory barrier.. > atomic_cmpset_64() ends with a memory barrier. So isn't the > memory barrier in atomic_cmpset_acq_64() extranious? Eg, you have 2 > memory barriers back to back. The one in atomic_cmpset_64 is obsolete. atomic_cmpset_acq_64 is correct. The problem I have with removing the barrier in atomic_cmpset_64 and the like is that it makes an additional difference to i386 and I think we already have enough alpha only bugs to build a new set just for a bit performance. > At any rate, the overhead just plain stinks. At nearly 0.5 us per > mutex, they are just way too expensive. > > > You even need them on UP systems to enshure instruction order. > > Alphas older than ev6 are in-order processors with regard to > instruction order, so presumably they are not needed on ev56 and older > machines? Sounds logical. But if you want to implement specialized tuning in regards of CPU and UP, then there is much more that can be done. Terry has brought up many interessting ideas of what can be done if you know your machine. > > Some barriers can be removed in other atomic functions or replaced > > with write barriers. > > Currently our alpha_wmb() is mappend to mb, which is done because it's > > also done in NetBSD - possibly as a workaround for a unnamed problem. > > > > > FWIW, The appended diff to remove them reducess null system call > > > latency to 1.6us with kern.giant.proc=1, and 1.4us with > > > kern.giant.proc=0. I'm about to start a buildworld with it, but I > > > don't have any SMP boxes. > > > > What kind of CPU is in your XP1000? > > 21264. > > > On >= 21164 the CPU can be initialised into UP operation so that mb > > and wmb fall back to just an contraint on instruction order. > > Also locked instructions can be handled inside the CPU in that case. > > Please elaborate. How does this work with respect to devices? That > sounds quite scary for the atomic code in general. What I wrote was only the short story for the intended use of the atomic functions. Memory barries still have their normal semantics in respect to other memory regions so if you write data into memory, do a barrier and tell the device to dma it, then the barrier still works because the device register is marked uncacheable. But barriers shouldn't be very expensive, what makes them expensive is the required combination with locked instuctions. > FWIW, my machine survived the buildworld. The removal of the memory > barriers chopped nearly 7 minutes off the time: There is not a big chance to loose the insturction ordering race I guess. > with MB: > 8699.25 real 6985.64 user 1379.72 sys > > without MB: > 8298.44 real 7010.03 user 969.02 sys Would be insteresting what happens if you remove only the obsolete barriers. What is required is the following: ... /* acq */ atomic (locked) memory access mb ... /* rel */ mb atomic (locked) memory access ... But what we currently have is ... /* acq */ atomic (locked) memory access mb mb ... /* rel */ mb atomic (locked) memory access mb ... -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 7:35:52 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D157937B400; Thu, 5 Sep 2002 07:35:45 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6B0B143E65; Thu, 5 Sep 2002 07:35:41 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id KAA25130; Thu, 5 Sep 2002 10:35:40 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85EZAP58186; Thu, 5 Sep 2002 10:35:10 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.27550.588643.958037@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 10:35:10 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: <20020905084824.GJ87724@cicely5.cicely.de> References: <15734.29725.515274.183629@grasshopper.cs.duke.edu> <20020904223255.GI87724@cicely5.cicely.de> <15734.39976.326598.176742@grasshopper.cs.duke.edu> <20020905084824.GJ87724@cicely5.cicely.de> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter writes: > > with MB: > > 8699.25 real 6985.64 user 1379.72 sys > > > > without MB: > > 8298.44 real 7010.03 user 969.02 sys > > Would be insteresting what happens if you remove only the obsolete > barriers. The lmbench perf is just the same as with the wrongheaded way I was doing it earlier -- 1.6us with kern.giant.proc=1, and 1.4us with kern.giant.proc=0. To refresh, it is 2.0/1.6 with the stock code. I'm starting a buildworld now. > What is required is the following: > ... > /* acq */ > atomic (locked) memory access > mb > ... > /* rel */ > mb > atomic (locked) memory access > ... > > But what we currently have is > ... > /* acq */ > atomic (locked) memory access > mb > mb > ... > /* rel */ > mb > atomic (locked) memory access > mb I'm using the appended diff, which I think is what you're suggesting: Cheers, Drew Index: atomic.h =================================================================== RCS file: /home/ncvs/src/sys/alpha/include/atomic.h,v retrieving revision 1.14 diff -u -r1.14 atomic.h --- atomic.h 17 May 2002 05:45:39 -0000 1.14 +++ atomic.h 5 Sep 2002 14:08:50 -0000 @@ -416,35 +416,61 @@ static __inline u_int32_t atomic_cmpset_acq_32(volatile u_int32_t *p, u_int32_t cmpval, u_int32_t newval) { - int retval; - - retval = atomic_cmpset_32(p, cmpval, newval); - alpha_mb(); - return (retval); + return (atomic_cmpset_32(p, cmpval, newval)); } static __inline u_int32_t atomic_cmpset_rel_32(volatile u_int32_t *p, u_int32_t cmpval, u_int32_t newval) { + u_int32_t ret; + alpha_mb(); - return (atomic_cmpset_32(p, cmpval, newval)); + __asm __volatile ( + "1:\tldl_l %0, %4\n\t" /* load old value */ + "cmpeq %0, %2, %0\n\t" /* compare */ + "beq %0, 2f\n\t" /* exit if not equal */ + "mov %3, %0\n\t" /* value to store */ + "stl_c %0, %1\n\t" /* attempt to store */ + "beq %0, 3f\n\t" /* if it failed, spin */ + "2:\n" /* done */ + ".section .text3,\"ax\"\n" /* improve branch prediction */ + "3:\tbr 1b\n" /* try again */ + ".previous\n" + : "=&r" (ret), "=m" (*p) + : "r" ((long)(int)cmpval), "r" (newval), "m" (*p) + : "memory"); + + return ret; } static __inline u_int64_t atomic_cmpset_acq_64(volatile u_int64_t *p, u_int64_t cmpval, u_int64_t newval) { - int retval; - - retval = atomic_cmpset_64(p, cmpval, newval); - alpha_mb(); - return (retval); + return (atomic_cmpset_64(p, cmpval, newval)); } static __inline u_int64_t atomic_cmpset_rel_64(volatile u_int64_t *p, u_int64_t cmpval, u_int64_t newval) { + u_int64_t ret; + alpha_mb(); - return (atomic_cmpset_64(p, cmpval, newval)); + __asm __volatile ( + "1:\tldq_l %0, %4\n\t" /* load old value */ + "cmpeq %0, %2, %0\n\t" /* compare */ + "beq %0, 2f\n\t" /* exit if not equal */ + "mov %3, %0\n\t" /* value to store */ + "stq_c %0, %1\n\t" /* attempt to store */ + "beq %0, 3f\n\t" /* if it failed, spin */ + "2:\n" /* done */ + ".section .text3,\"ax\"\n" /* improve branch prediction */ + "3:\tbr 1b\n" /* try again */ + ".previous\n" + : "=&r" (ret), "=m" (*p) + : "r" (cmpval), "r" (newval), "m" (*p) + : "memory"); + + return ret; } #define atomic_cmpset_acq_int atomic_cmpset_acq_32 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 7:43:34 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id AA9C737B40A for ; Thu, 5 Sep 2002 07:43:31 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id 16D5143E65 for ; Thu, 5 Sep 2002 07:43:30 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 26340 invoked from network); 5 Sep 2002 14:43:27 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 14:43:27 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85EhRBv011235; Thu, 5 Sep 2002 10:43:27 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.27550.588643.958037@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 10:43:27 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: alpha performance on -current Cc: freebsd-alpha@FreeBSD.ORG, ticso@cicely.de Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > I'm using the appended diff, which I think is what you're suggesting: I think what he is really suggesting is to remove the mb's from all the atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel versions should be wrappers that add mb's. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 7:52: 5 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 41C1A37B400; Thu, 5 Sep 2002 07:52:03 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5A36E43E3B; Thu, 5 Sep 2002 07:52:02 -0700 (PDT) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g85Epw6i070911 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 16:51:59 +0200 (CEST) (envelope-from ticso@cicely5.cicely.de) Received: from cicely5.cicely.de (localhost [IPv6:::1]) by cicely5.cicely.de (8.12.1/8.12.1) with ESMTP id g85EpuUV094327 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 16:51:57 +0200 (CEST)?g (envelope-from ticso@cicely5.cicely.de) Received: (from ticso@localhost) by cicely5.cicely.de (8.12.1/8.12.1/Submit) id g85Epub1094326; Thu, 5 Sep 2002 16:51:56 +0200 (CEST)?g (envelope-from ticso) Date: Thu, 5 Sep 2002 16:51:56 +0200 From: Bernd Walter To: John Baldwin Cc: Andrew Gallatin , freebsd-alpha@FreeBSD.org, ticso@cicely.de Subject: Re: alpha performance on -current Message-ID: <20020905145156.GB94194@cicely5.cicely.de> Reply-To: ticso@cicely.de References: <15735.27550.588643.958037@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Operating-System: FreeBSD cicely5.cicely.de 5.0-CURRENT i386 User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 10:43:27AM -0400, John Baldwin wrote: > > On 05-Sep-2002 Andrew Gallatin wrote: > > I'm using the appended diff, which I think is what you're suggesting: > > I think what he is really suggesting is to remove the mb's from all the > atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel > versions should be wrappers that add mb's. Exactly. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 8:10:43 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3131A37B401; Thu, 5 Sep 2002 08:10:42 -0700 (PDT) Received: from smtp.web.de (smtp02.web.de [217.72.192.151]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4E49843E3B; Thu, 5 Sep 2002 08:10:41 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [217.225.196.147] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17myGl-0000ui-00; Thu, 05 Sep 2002 17:10:07 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id E56EB3EB; Thu, 5 Sep 2002 17:10:04 +0200 (CEST) Received: from jan-linux.lan (jan-linux.lan [192.168.0.20]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 934A2126; Thu, 5 Sep 2002 17:09:45 +0200 (CEST) Subject: Re: alpha performance on -current From: Jan Lentfer To: ticso@cicely.de Cc: John Baldwin , Andrew Gallatin , freebsd-alpha@FreeBSD.org In-Reply-To: <20020905145156.GB94194@cicely5.cicely.de> References: <15735.27550.588643.958037@grasshopper.cs.duke.edu> <20020905145156.GB94194@cicely5.cicely.de> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.7-1mdk Date: 05 Sep 2002 17:09:43 +0200 Message-Id: <1031238584.2988.14.camel@jan-linux.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Am Don, 2002-09-05 um 16.51 schrieb Bernd Walter: > On Thu, Sep 05, 2002 at 10:43:27AM -0400, John Baldwin wrote: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > I'm using the appended diff, which I think is what you're suggesting: > > > > I think what he is really suggesting is to remove the mb's from all the > > atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel > > versions should be wrappers that add mb's. > > Exactly. OK, I didn't get anything from the first sentence until now. What the heck are you guys talking about??? I mean it, could someone explain or even better point me to some resources so I can actually understand what you guys mean? This is about kernel design and memory managment, right? Any recommended readings? Thanks a lot in advance, Jan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 8:22:24 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 358CE37B400 for ; Thu, 5 Sep 2002 08:22:22 -0700 (PDT) Received: from mail.speakeasy.net (mail16.speakeasy.net [216.254.0.216]) by mx1.FreeBSD.org (Postfix) with ESMTP id C005743E75 for ; Thu, 5 Sep 2002 08:22:21 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 3516 invoked from network); 5 Sep 2002 15:22:58 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail16.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 15:22:58 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85FMJBv011339; Thu, 5 Sep 2002 11:22:20 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <1031238584.2988.14.camel@jan-linux.lan> Date: Thu, 05 Sep 2002 11:22:19 -0400 (EDT) From: John Baldwin To: Jan Lentfer Subject: Re: alpha performance on -current Cc: freebsd-alpha@FreeBSD.org, Andrew Gallatin , ticso@cicely.de Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Jan Lentfer wrote: > Am Don, 2002-09-05 um 16.51 schrieb Bernd Walter: >> On Thu, Sep 05, 2002 at 10:43:27AM -0400, John Baldwin wrote: >> > >> > On 05-Sep-2002 Andrew Gallatin wrote: >> > > I'm using the appended diff, which I think is what you're suggesting: >> > >> > I think what he is really suggesting is to remove the mb's from all the >> > atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel >> > versions should be wrappers that add mb's. >> >> Exactly. > > OK, I didn't get anything from the first sentence until now. What the > heck are you guys talking about??? I mean it, could someone explain or > even better point me to some resources so I can actually understand what > you guys mean? This is about kernel design and memory managment, right? > Any recommended readings? > > Thanks a lot in advance, This is about implementing atomic operations on Alpha CPU's for use by the kernel. You can read the atomic(9) man page on current for more info about this. An atomic operation just guarantees that the operation it is performing is atomic, hence the fact that normal atomic_foo() oops don't need memory barriers. Only the versions of atomic ops (such as atomic_foo_acq() or atomic_foo_rel()) require that memory barriers be used to guarantee specific ordering restraints. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 8:42:36 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C155F37B400; Thu, 5 Sep 2002 08:42:33 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id D818943E42; Thu, 5 Sep 2002 08:42:32 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id LAA27928; Thu, 5 Sep 2002 11:42:27 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85Ffvw58255; Thu, 5 Sep 2002 11:41:57 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.31557.85006.684144@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 11:41:57 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: <20020905145156.GB94194@cicely5.cicely.de> References: <15735.27550.588643.958037@grasshopper.cs.duke.edu> <20020905145156.GB94194@cicely5.cicely.de> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter writes: > On Thu, Sep 05, 2002 at 10:43:27AM -0400, John Baldwin wrote: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > I'm using the appended diff, which I think is what you're suggesting: > > > > I think what he is really suggesting is to remove the mb's from all the > > atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel > > versions should be wrappers that add mb's. > > Exactly. OK, I'm bound to make a hash of things. Can you show me what you mean? Thanks, Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 9:10:54 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8435E37B400 for ; Thu, 5 Sep 2002 09:10:52 -0700 (PDT) Received: from qescan1.qgraph.com (QESCAN1.qgraph.com [206.158.124.7]) by mx1.FreeBSD.org (Postfix) with SMTP id 4FA5443E75 for ; Thu, 5 Sep 2002 09:10:51 -0700 (PDT) (envelope-from Aaron.Schroeder@qg.com) Received: from 192.168.200.28 by qescan1.qgraph.com (InterScan E-Mail VirusWall NT); Thu, 05 Sep 2002 11:10:50 -0500 Received: by sxsmtp1.qgraph.com with Internet Mail Service (5.5.2656.59) id ; Thu, 5 Sep 2002 11:11:54 -0500 Message-ID: <025596A38A894B45AFE62346A6BF47464E2C12@waexch1.qgraph.com> From: "Schroeder, Aaron" To: "'freebsd-alpha@freebsd.org'" Subject: MT possibly broken? Date: Thu, 5 Sep 2002 11:10:44 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2656.59) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Hello all, I have a PWS500au running 4.6-STABLE with a DEC TZL09 DAT tape drive. I was seeing some problems with this drive when it was hooked onto the same SCSI chain as my hard drives, actually, I saw problems when it was hooked up to the same Qlogic card. I was getting I/O errors when trying to write to and erase tapes. I was using 'mt' at first, then I installed flexbackup from ports. Flexbackup uses 'mt' commands, so that did not work for me as well. So, I installed a SymBios PCI SCSI card into my machine and hung the tape drive off of it, by itself. I ran through the same procedures as above, I used 'mt' to rewind and erase tapes, rather 'mt' attempted to erase the tapes, I still got I/O errors. Flexbackup, again, was the same story. So, I read up on the 'dump' and 'restore' utilities and decided to try those out. I did a 'dump' of my /home folder and it said that it successfully dumped to tape! Interesting, so I tried to use 'restore' to see if 'dump' did what it said it did. So, I see that I have my /home folder on the tape, I select a couple files to extract the from tape, again, it worked! I am just happy that I can back up my system at this point. So, just to confirm, I ran 'mt -f /dev/sa0 erase' and I received a I/O error. Unless I am doing something wrong, it seems to me that 'mt' may have an issue. Be it either that it is broken, or it doesn't like my tape drive and dump does. I do have a TZ89 DLT drive that I can try as well, if anyone thinks that is needed. Thanks, AJ Schroeder To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 9:27:41 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EB7CA37B405 for ; Thu, 5 Sep 2002 09:27:36 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id 84CC243E4A for ; Thu, 5 Sep 2002 09:27:36 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 311 invoked from network); 5 Sep 2002 16:27:34 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 16:27:34 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85GRYBv011534; Thu, 5 Sep 2002 12:27:34 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.31557.85006.684144@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 12:27:34 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: alpha performance on -current Cc: freebsd-alpha@FreeBSD.ORG, ticso@cicely.de Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org :q On 05-Sep-2002 Andrew Gallatin wrote: > > Bernd Walter writes: > > On Thu, Sep 05, 2002 at 10:43:27AM -0400, John Baldwin wrote: > > > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > I'm using the appended diff, which I think is what you're suggesting: > > > > > > I think what he is really suggesting is to remove the mb's from all the > > > atomic_foo ops that don't have _acq or _rel in them. The _acq and _rel > > > versions should be wrappers that add mb's. > > > > Exactly. > > OK, I'm bound to make a hash of things. Can you show me what you > mean? Here's an example from just one of the atomic ops: --- atomic.h 17 May 2002 05:45:39 -0000 1.14 +++ atomic.h 5 Sep 2002 16:24:28 -0000 @@ -51,16 +51,15 @@ u_int32_t temp; __asm __volatile ( "1:\tldl_l %0, %2\n\t" /* load old value */ "bis %0, %3, %0\n\t" /* calculate new value */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" : "=&r" (temp), "=m" (*p) : "m" (*p), "r" (v) : "memory"); } Thus, the only atomic ops that would have any barriers in them are the _acq and _rel variants. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 10:15:56 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id A433E37B400; Thu, 5 Sep 2002 10:15:55 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id B57BB43E6A; Thu, 5 Sep 2002 10:15:54 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id NAA01855; Thu, 5 Sep 2002 13:15:53 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85HFNa58335; Thu, 5 Sep 2002 13:15:23 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.37163.753487.570198@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 13:15:23 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.ORG, ticso@cicely.de Subject: Re: alpha performance on -current In-Reply-To: References: <15735.31557.85006.684144@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > OK, I'm bound to make a hash of things. Can you show me what you > > mean? > > Here's an example from just one of the atomic ops: OK, that's what I thought you said. I'm trying it now. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 10:35:11 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2148937B401; Thu, 5 Sep 2002 10:35:10 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id D6D5043E65; Thu, 5 Sep 2002 10:35:08 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id NAA02783; Thu, 5 Sep 2002 13:35:08 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85HYcC58349; Thu, 5 Sep 2002 13:34:38 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.38318.66517.366474@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 13:34:38 -0400 (EDT) To: jhb@freebsd.org Cc: freebsd-alpha@freebsd.org Subject: ithread preemption X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org I've forgotten -- What are the symptoms of ithread preemption causing troubles on alpha? I have one (probably dumb) idea: Is the ithread preemption code guaranteed to switch back to the preempted thread when the ithread completes or blocks? And continue through to the end of the interrupt dispatch code, returning back to the palcode? Eg, I'm wondering if it could it have anything to do with continuing to run at IPL_IO when we shouldn't be (like running a normal process at that priority that the scheduler happened to run by mistake). Thanks, Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 12: 6:45 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D995D37B400 for ; Thu, 5 Sep 2002 12:06:42 -0700 (PDT) Received: from mail.speakeasy.net (mail15.speakeasy.net [216.254.0.215]) by mx1.FreeBSD.org (Postfix) with ESMTP id 79CAA43E65 for ; Thu, 5 Sep 2002 12:06:42 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 19780 invoked from network); 5 Sep 2002 19:06:41 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail15.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 19:06:41 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85J6eBv012091; Thu, 5 Sep 2002 15:06:40 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.38318.66517.366474@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 15:06:40 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: RE: ithread preemption Cc: freebsd-alpha@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > I've forgotten -- What are the symptoms of ithread preemption causing > troubles on alpha? Hangs on SMP under load. > I have one (probably dumb) idea: Is the ithread preemption code > guaranteed to switch back to the preempted thread when the ithread > completes or blocks? And continue through to the end of the interrupt > dispatch code, returning back to the palcode? It is not guaranteed to do that. > Eg, I'm wondering if it could it have anything to do with continuing > to run at IPL_IO when we shouldn't be (like running a normal > process at that priority that the scheduler happened to run by mistake). Nope, IPL changes when you switch threads. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 12:20:56 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id BF3A937B401; Thu, 5 Sep 2002 12:20:53 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 77C9843E7B; Thu, 5 Sep 2002 12:20:51 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id PAA06484; Thu, 5 Sep 2002 15:20:50 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85JKKu58439; Thu, 5 Sep 2002 15:20:20 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.44660.835003.901974@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 15:20:20 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.org Subject: RE: ithread preemption In-Reply-To: References: <15735.38318.66517.366474@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > I've forgotten -- What are the symptoms of ithread preemption causing > > troubles on alpha? > > Hangs on SMP under load. > > > I have one (probably dumb) idea: Is the ithread preemption code > > guaranteed to switch back to the preempted thread when the ithread > > completes or blocks? And continue through to the end of the interrupt > > dispatch code, returning back to the palcode? > > It is not guaranteed to do that. What keeps you from (eventually) running out of kernel stack space then, as the interrupts keep coming in? Eg, would it help to keep a per-cpu (not per thread) interrupt nesting count and refuse to do preemption if the interrupt nesting count exceeds a small constant? Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 12:35:39 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DBD2E37B400 for ; Thu, 5 Sep 2002 12:35:36 -0700 (PDT) Received: from mail.speakeasy.net (mail15.speakeasy.net [216.254.0.215]) by mx1.FreeBSD.org (Postfix) with ESMTP id 74BA743E75 for ; Thu, 5 Sep 2002 12:35:36 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 4282 invoked from network); 5 Sep 2002 19:35:35 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail15.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 19:35:35 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85JZXBv012169; Thu, 5 Sep 2002 15:35:34 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.44660.835003.901974@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 15:35:33 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: RE: ithread preemption Cc: freebsd-alpha@FreeBSD.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > I've forgotten -- What are the symptoms of ithread preemption causing > > > troubles on alpha? > > > > Hangs on SMP under load. > > > > > I have one (probably dumb) idea: Is the ithread preemption code > > > guaranteed to switch back to the preempted thread when the ithread > > > completes or blocks? And continue through to the end of the interrupt > > > dispatch code, returning back to the palcode? > > > > It is not guaranteed to do that. > > What keeps you from (eventually) running out of kernel stack space > then, as the interrupts keep coming in? The thread that received the interrupt stays at the high IPL until it returns. When you switch to another thread you are on another stack and you can take an interrupt ok. When we switch back to an interrupted thread, it executes at the raised IPL until it returns back to the PAL code. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13: 3:19 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id BEFC237B400; Thu, 5 Sep 2002 13:03:16 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id D1CAF43E7B; Thu, 5 Sep 2002 13:03:15 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA08226; Thu, 5 Sep 2002 16:03:15 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85K2ip58532; Thu, 5 Sep 2002 16:02:44 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.47204.905352.900631@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:02:44 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.org Subject: RE: ithread preemption In-Reply-To: References: <15735.44660.835003.901974@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > John Baldwin writes: > > > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > > > I've forgotten -- What are the symptoms of ithread preemption causing > > > > troubles on alpha? > > > > > > Hangs on SMP under load. > > > > > > > I have one (probably dumb) idea: Is the ithread preemption code > > > > guaranteed to switch back to the preempted thread when the ithread > > > > completes or blocks? And continue through to the end of the interrupt > > > > dispatch code, returning back to the palcode? > > > > > > It is not guaranteed to do that. > > > > What keeps you from (eventually) running out of kernel stack space > > then, as the interrupts keep coming in? > > The thread that received the interrupt stays at the high IPL until it > returns. When you switch to another thread you are on another stack > and you can take an interrupt ok. When we switch back to an interrupted > thread, it executes at the raised IPL until it returns back to the PAL > code. OK, so the interrupted thread will (eventually) return back to PAL. But, theoritically, under heavy load we could have lots of threads preempted. And lots of interrupts pending which never returned to PAL. Are we certain that this doesn't somehow violate assumptions made by pal? Does any other OS work like this? Perhaps we should only allow a small number of preemptions at any one time. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13: 7: 6 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3C03237B400 for ; Thu, 5 Sep 2002 13:07:03 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id C6C1F43E72 for ; Thu, 5 Sep 2002 13:07:02 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 7101 invoked from network); 5 Sep 2002 20:07:00 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 20:07:00 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85K70Bv012279; Thu, 5 Sep 2002 16:07:00 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.47204.905352.900631@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 16:07:00 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: RE: ithread preemption Cc: freebsd-alpha@FreeBSD.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > John Baldwin writes: > > > > > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > > > > > I've forgotten -- What are the symptoms of ithread preemption causing > > > > > troubles on alpha? > > > > > > > > Hangs on SMP under load. > > > > > > > > > I have one (probably dumb) idea: Is the ithread preemption code > > > > > guaranteed to switch back to the preempted thread when the ithread > > > > > completes or blocks? And continue through to the end of the interrupt > > > > > dispatch code, returning back to the palcode? > > > > > > > > It is not guaranteed to do that. > > > > > > What keeps you from (eventually) running out of kernel stack space > > > then, as the interrupts keep coming in? > > > > The thread that received the interrupt stays at the high IPL until it > > returns. When you switch to another thread you are on another stack > > and you can take an interrupt ok. When we switch back to an interrupted > > thread, it executes at the raised IPL until it returns back to the PAL > > code. > > OK, so the interrupted thread will (eventually) return back to PAL. > > But, theoritically, under heavy load we could have lots of threads > preempted. And lots of interrupts pending which never returned to > PAL. Are we certain that this doesn't somehow violate assumptions made > by pal? Does any other OS work like this? Solaris doesn't run on alpha, but it also a bit different in its approach. I do wonder if there is a way we can violate an assumption in PAL due to migration though. That is, a thread could return to PAL on a different CPU than the one the interrupt was originally sent to. This might explain why only SMP has problems. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:12:29 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D71A137B400; Thu, 5 Sep 2002 13:12:27 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 035CF43E3B; Thu, 5 Sep 2002 13:12:27 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA08581; Thu, 5 Sep 2002 16:12:26 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85KBuo58557; Thu, 5 Sep 2002 16:11:56 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.47756.501169.199225@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:11:56 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.ORG Subject: RE: ithread preemption In-Reply-To: References: <15735.47204.905352.900631@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > Solaris doesn't run on alpha, but it also a bit different in its approach. > I do wonder if there is a way we can violate an assumption in PAL due to > migration though. That is, a thread could return to PAL on a different > CPU than the one the interrupt was originally sent to. This might explain > why only SMP has problems. > Hey ... I think you have it on the nose! That makes the most sense I've heard yet. Do we have any way to bind a thread to a cpu? Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:13:32 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 468A737B400; Thu, 5 Sep 2002 13:13:25 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3396D43E72; Thu, 5 Sep 2002 13:13:24 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA08598; Thu, 5 Sep 2002 16:13:23 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85KCrM58560; Thu, 5 Sep 2002 16:12:53 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.47813.266130.216534@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:12:53 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.org, ticso@cicely.de Subject: Re: alpha performance on -current In-Reply-To: References: <15735.31557.85006.684144@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org OK, if the appended is what you had in mind, then I'm a much happier camper. This patch: 8314.06 real 7025.32 user 956.14 sys Stock kernel: 8699.25 real 6985.64 user 1379.72 sys A 30% reduction in system time is pretty good. Cheers, Drew Index: include/atomic.h =================================================================== RCS file: /home/ncvs/src/sys/alpha/include/atomic.h,v retrieving revision 1.14 diff -u -r1.14 atomic.h --- include/atomic.h 17 May 2002 05:45:39 -0000 1.14 +++ include/atomic.h 5 Sep 2002 17:03:21 -0000 @@ -55,7 +55,6 @@ "bis %0, %3, %0\n\t" /* calculate new value */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -73,7 +72,6 @@ "bic %0, %3, %0\n\t" /* calculate new value */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -91,7 +89,6 @@ "addl %0, %3, %0\n\t" /* calculate new value */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -109,7 +106,6 @@ "subl %0, %3, %0\n\t" /* calculate new value */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -130,7 +126,7 @@ "beq %1,2f\n\t" /* if the store failed, spin */ "br 3f\n" /* it worked, exit */ "2:\tbr 1b\n" /* *addr not updated, loop */ - "3:\tmb\n" /* it worked */ + "3:\n" /* it worked */ : "=&r"(result), "=&r"(temp), "=m" (*addr) : "m"(*addr) : "memory"); @@ -147,7 +143,6 @@ "bis %0, %3, %0\n\t" /* calculate new value */ "stq_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -165,7 +160,6 @@ "bic %0, %3, %0\n\t" /* calculate new value */ "stq_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -183,7 +177,6 @@ "addq %0, %3, %0\n\t" /* calculate new value */ "stq_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -201,7 +194,6 @@ "subq %0, %3, %0\n\t" /* calculate new value */ "stq_c %0, %1\n\t" /* attempt to store */ "beq %0, 2f\n\t" /* spin if failed */ - "mb\n\t" /* drain to memory */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "2:\tbr 1b\n" /* try again */ ".previous\n" @@ -222,7 +214,7 @@ "beq %1,2f\n\t" /* if the store failed, spin */ "br 3f\n" /* it worked, exit */ "2:\tbr 1b\n" /* *addr not updated, loop */ - "3:\tmb\n" /* it worked */ + "3:\n" /* it worked */ : "=&r"(result), "=&r"(temp), "=m" (*addr) : "m"(*addr) : "memory"); @@ -257,7 +249,7 @@ atomic_##NAME##_acq_##WIDTH(volatile u_int##WIDTH##_t *p, u_int##WIDTH##_t v)\ { \ atomic_##NAME##_##WIDTH(p, v); \ - /* alpha_mb(); */ \ + alpha_mb(); \ } \ \ static __inline void \ @@ -271,7 +263,7 @@ atomic_##NAME##_acq_##TYPE(volatile u_int##WIDTH##_t *p, u_int##WIDTH##_t v)\ { \ atomic_##NAME##_##WIDTH(p, v); \ - /* alpha_mb(); */ \ + alpha_mb(); \ } \ \ static __inline void \ @@ -361,7 +353,6 @@ "mov %3, %0\n\t" /* value to store */ "stl_c %0, %1\n\t" /* attempt to store */ "beq %0, 3f\n\t" /* if it failed, spin */ - "mb\n\t" /* drain to memory */ "2:\n" /* done */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "3:\tbr 1b\n" /* try again */ @@ -390,7 +381,6 @@ "mov %3, %0\n\t" /* value to store */ "stq_c %0, %1\n\t" /* attempt to store */ "beq %0, 3f\n\t" /* if it failed, spin */ - "mb\n\t" /* drain to memory */ "2:\n" /* done */ ".section .text3,\"ax\"\n" /* improve branch prediction */ "3:\tbr 1b\n" /* try again */ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:15:26 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3F0B037B400; Thu, 5 Sep 2002 13:15:23 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id CF86043E42; Thu, 5 Sep 2002 13:15:21 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g85KFB6K074031 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 22:15:12 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g85KFANk002149 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 22:15:10 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g85KEhD3013214; Thu, 5 Sep 2002 22:14:44 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g85KEh4t013213; Thu, 5 Sep 2002 22:14:43 +0200 (CEST) Date: Thu, 5 Sep 2002 22:14:43 +0200 From: Bernd Walter To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption Message-ID: <20020905201443.GD13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <15735.44660.835003.901974@grasshopper.cs.duke.edu> <15735.47204.905352.900631@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15735.47204.905352.900631@grasshopper.cs.duke.edu> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 04:02:44PM -0400, Andrew Gallatin wrote: > > John Baldwin writes: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > John Baldwin writes: > > > > > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > > > > > I've forgotten -- What are the symptoms of ithread preemption causing > > > > > troubles on alpha? > > > > > > > > Hangs on SMP under load. > > > > > > > > > I have one (probably dumb) idea: Is the ithread preemption code > > > > > guaranteed to switch back to the preempted thread when the ithread > > > > > completes or blocks? And continue through to the end of the interrupt > > > > > dispatch code, returning back to the palcode? > > > > > > > > It is not guaranteed to do that. > > > > > > What keeps you from (eventually) running out of kernel stack space > > > then, as the interrupts keep coming in? > > > > The thread that received the interrupt stays at the high IPL until it > > returns. When you switch to another thread you are on another stack > > and you can take an interrupt ok. When we switch back to an interrupted > > thread, it executes at the raised IPL until it returns back to the PAL > > code. > > OK, so the interrupted thread will (eventually) return back to PAL. OK - I have some basic understandig problems here. Why should ithreads ever return to PAL? Why is IPL raised while an ithread is running? From what I understood before the interrupt handler, which is called from PAL, just triggers the ithread, block the intline and returns. > But, theoritically, under heavy load we could have lots of threads > preempted. And lots of interrupts pending which never returned to > PAL. Are we certain that this doesn't somehow violate assumptions made > by pal? Does any other OS work like this? > > Perhaps we should only allow a small number of preemptions at any one > time. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:21:14 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7763137B400; Thu, 5 Sep 2002 13:21:12 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id A416E43E65; Thu, 5 Sep 2002 13:21:11 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA08962; Thu, 5 Sep 2002 16:21:11 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85KKfe58577; Thu, 5 Sep 2002 16:20:41 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.48281.30915.800894@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:20:41 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption In-Reply-To: <20020905201443.GD13050@cicely9.cicely.de> References: <15735.44660.835003.901974@grasshopper.cs.duke.edu> <15735.47204.905352.900631@grasshopper.cs.duke.edu> <20020905201443.GD13050@cicely9.cicely.de> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter writes: > OK - I have some basic understandig problems here. > > Why should ithreads ever return to PAL? Because PAL initiates the interrupt and creates a stackframe on the interrupted thread's kernel stack. It then calls the OSes interrupt vector. In order to restore the state of the world to be like it was before the interrupt happened, we need to return back to pal. > Why is IPL raised while an ithread is running? Pal raises the IPL to the appropriate level to block further interrupts from the same source before calling into the OS. > >From what I understood before the interrupt handler, which is called > from PAL, just triggers the ithread, block the intline and returns. > The -stable code is easier to understand.. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:31: 7 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EE5B737B401; Thu, 5 Sep 2002 13:31:04 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6168843E6A; Thu, 5 Sep 2002 13:31:03 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g85KUw6K074213 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 22:30:59 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g85KUuNk002232 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 22:30:56 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g85KUUD3013246; Thu, 5 Sep 2002 22:30:30 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g85KUTFQ013245; Thu, 5 Sep 2002 22:30:29 +0200 (CEST) Date: Thu, 5 Sep 2002 22:30:29 +0200 From: Bernd Walter To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG, ticso@cicely.de Subject: Re: alpha performance on -current Message-ID: <20020905203029.GE13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <15735.31557.85006.684144@grasshopper.cs.duke.edu> <15735.47813.266130.216534@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15735.47813.266130.216534@grasshopper.cs.duke.edu> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 04:12:53PM -0400, Andrew Gallatin wrote: > > OK, if the appended is what you had in mind, then I'm a much happier > camper. Yes. > This patch: > > 8314.06 real 7025.32 user 956.14 sys > > Stock kernel: > > 8699.25 real 6985.64 user 1379.72 sys > > > A 30% reduction in system time is pretty good. I never thought that it would make such a big difference. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:31:10 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 830E837B400 for ; Thu, 5 Sep 2002 13:31:06 -0700 (PDT) Received: from mail.speakeasy.net (mail16.speakeasy.net [216.254.0.216]) by mx1.FreeBSD.org (Postfix) with ESMTP id D76D843E65 for ; Thu, 5 Sep 2002 13:31:05 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 28648 invoked from network); 5 Sep 2002 20:31:42 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail16.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 20:31:42 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85KV3Bv012377; Thu, 5 Sep 2002 16:31:03 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.47813.266130.216534@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 16:31:02 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: alpha performance on -current Cc: ticso@cicely.de, freebsd-alpha@FreeBSD.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > OK, if the appended is what you had in mind, then I'm a much happier > camper. Yep. > This patch: > > 8314.06 real 7025.32 user 956.14 sys > > Stock kernel: > > 8699.25 real 6985.64 user 1379.72 sys > > > A 30% reduction in system time is pretty good. Cool. I'll be sure and try to test it on the DS20 I have to make sure it is safe. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:31:14 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 66DDF37B401 for ; Thu, 5 Sep 2002 13:31:08 -0700 (PDT) Received: from mail.speakeasy.net (mail11.speakeasy.net [216.254.0.211]) by mx1.FreeBSD.org (Postfix) with ESMTP id BCF8743E65 for ; Thu, 5 Sep 2002 13:31:07 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 22669 invoked from network); 5 Sep 2002 20:31:06 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail11.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 20:31:06 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85KV5Bv012380; Thu, 5 Sep 2002 16:31:05 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.47756.501169.199225@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 16:31:05 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: RE: ithread preemption Cc: freebsd-alpha@FreeBSD.ORG Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > Solaris doesn't run on alpha, but it also a bit different in its approach. > > I do wonder if there is a way we can violate an assumption in PAL due to > > migration though. That is, a thread could return to PAL on a different > > CPU than the one the interrupt was originally sent to. This might explain > > why only SMP has problems. > > > > Hey ... I think you have it on the nose! That makes the most sense > I've heard yet. > > Do we have any way to bind a thread to a cpu? I used to. Then KSE3 was committed. I suppose I could rewrite it from scratch again. > Drew -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:31:18 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8915E37B400 for ; Thu, 5 Sep 2002 13:31:12 -0700 (PDT) Received: from mail.speakeasy.net (mail11.speakeasy.net [216.254.0.211]) by mx1.FreeBSD.org (Postfix) with ESMTP id C4F2D43E72 for ; Thu, 5 Sep 2002 13:31:09 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 22694 invoked from network); 5 Sep 2002 20:31:08 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail11.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 20:31:09 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85KV7Bv012383; Thu, 5 Sep 2002 16:31:07 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <20020905201443.GD13050@cicely9.cicely.de> Date: Thu, 05 Sep 2002 16:31:07 -0400 (EDT) From: John Baldwin To: ticso@cicely.de Subject: Re: ithread preemption Cc: freebsd-alpha@FreeBSD.ORG, Andrew Gallatin Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Bernd Walter wrote: > On Thu, Sep 05, 2002 at 04:02:44PM -0400, Andrew Gallatin wrote: >> >> John Baldwin writes: >> > >> > On 05-Sep-2002 Andrew Gallatin wrote: >> > > >> > > John Baldwin writes: >> > > > >> > > > On 05-Sep-2002 Andrew Gallatin wrote: >> > > > > >> > > > > I've forgotten -- What are the symptoms of ithread preemption causing >> > > > > troubles on alpha? >> > > > >> > > > Hangs on SMP under load. >> > > > >> > > > > I have one (probably dumb) idea: Is the ithread preemption code >> > > > > guaranteed to switch back to the preempted thread when the ithread >> > > > > completes or blocks? And continue through to the end of the interrupt >> > > > > dispatch code, returning back to the palcode? >> > > > >> > > > It is not guaranteed to do that. >> > > >> > > What keeps you from (eventually) running out of kernel stack space >> > > then, as the interrupts keep coming in? >> > >> > The thread that received the interrupt stays at the high IPL until it >> > returns. When you switch to another thread you are on another stack >> > and you can take an interrupt ok. When we switch back to an interrupted >> > thread, it executes at the raised IPL until it returns back to the PAL >> > code. >> >> OK, so the interrupted thread will (eventually) return back to PAL. > > OK - I have some basic understandig problems here. > > Why should ithreads ever return to PAL? > Why is IPL raised while an ithread is running? > > From what I understood before the interrupt handler, which is called > from PAL, just triggers the ithread, block the intline and returns. Yes, but when you throw preemption into the mix you can switch to the actual ithread when you schedule it, and you don't return to the original interrupted thread until later. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:41:21 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3731B37B400; Thu, 5 Sep 2002 13:41:18 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8790A43E42; Thu, 5 Sep 2002 13:41:17 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA09803; Thu, 5 Sep 2002 16:41:17 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85KekH58601; Thu, 5 Sep 2002 16:40:46 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.49486.931825.65696@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:40:46 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.org Subject: RE: ithread preemption In-Reply-To: References: <15735.47756.501169.199225@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > John Baldwin writes: > > > Solaris doesn't run on alpha, but it also a bit different in its approach. > > > I do wonder if there is a way we can violate an assumption in PAL due to > > > migration though. That is, a thread could return to PAL on a different > > > CPU than the one the interrupt was originally sent to. This might explain > > > why only SMP has problems. > > > > > > > Hey ... I think you have it on the nose! That makes the most sense > > I've heard yet. > > > > Do we have any way to bind a thread to a cpu? > > I used to. Then KSE3 was committed. I suppose I could rewrite it from > scratch again. > Ugh. It would be great if you could. Even if it doesn't fix this, it might be generally useful. Thanks! Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:46:45 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F05FF37B635; Thu, 5 Sep 2002 13:46:33 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5A74E43E81; Thu, 5 Sep 2002 13:46:31 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g85KkR6K074404 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 22:46:28 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g85KkPNk002349 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 22:46:25 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g85KjxD3013305; Thu, 5 Sep 2002 22:45:59 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g85KjwNH013304; Thu, 5 Sep 2002 22:45:58 +0200 (CEST) Date: Thu, 5 Sep 2002 22:45:58 +0200 From: Bernd Walter To: Andrew Gallatin Cc: ticso@cicely.de, John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption Message-ID: <20020905204558.GF13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <15735.44660.835003.901974@grasshopper.cs.duke.edu> <15735.47204.905352.900631@grasshopper.cs.duke.edu> <20020905201443.GD13050@cicely9.cicely.de> <15735.48281.30915.800894@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15735.48281.30915.800894@grasshopper.cs.duke.edu> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 04:20:41PM -0400, Andrew Gallatin wrote: > > Bernd Walter writes: > > OK - I have some basic understandig problems here. > > > > Why should ithreads ever return to PAL? > > Because PAL initiates the interrupt and creates a stackframe on the > interrupted thread's kernel stack. It then calls the OSes interrupt > vector. In order to restore the state of the world to be like it was > before the interrupt happened, we need to return back to pal. That's clear, but this is not the ithread. ithreads never return - see ithread_loop() in kern_intr.c. > > Why is IPL raised while an ithread is running? > > Pal raises the IPL to the appropriate level to block further > interrupts from the same source before calling into the OS. > > > >From what I understood before the interrupt handler, which is called > > from PAL, just triggers the ithread, block the intline and returns. > > > > The -stable code is easier to understand.. As it was on the C64 =) -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:49:40 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 354F737B400; Thu, 5 Sep 2002 13:49:38 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 3AF8B43E75; Thu, 5 Sep 2002 13:49:37 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g85KnU6K074450 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Thu, 5 Sep 2002 22:49:33 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g85KnSNk002369 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Thu, 5 Sep 2002 22:49:29 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g85Kn2D3013312; Thu, 5 Sep 2002 22:49:02 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g85Kn2NT013311; Thu, 5 Sep 2002 22:49:02 +0200 (CEST) Date: Thu, 5 Sep 2002 22:49:02 +0200 From: Bernd Walter To: John Baldwin Cc: ticso@cicely.de, freebsd-alpha@FreeBSD.ORG, Andrew Gallatin Subject: Re: ithread preemption Message-ID: <20020905204901.GG13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <20020905201443.GD13050@cicely9.cicely.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 04:31:07PM -0400, John Baldwin wrote: > > On 05-Sep-2002 Bernd Walter wrote: > > OK - I have some basic understandig problems here. > > > > Why should ithreads ever return to PAL? > > Why is IPL raised while an ithread is running? > > > > From what I understood before the interrupt handler, which is called > > from PAL, just triggers the ithread, block the intline and returns. > > Yes, but when you throw preemption into the mix you can switch to the > actual ithread when you schedule it, and you don't return to the original > interrupted thread until later. Now I got it. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:50:36 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7BC0E37B406; Thu, 5 Sep 2002 13:50:29 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id D5BEB43E65; Thu, 5 Sep 2002 13:50:27 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA10144; Thu, 5 Sep 2002 16:50:27 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85Knvq58615; Thu, 5 Sep 2002 16:49:57 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.50037.249179.284948@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:49:57 -0400 (EDT) To: ticso@cicely.de Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption In-Reply-To: <20020905204558.GF13050@cicely9.cicely.de> References: <15735.44660.835003.901974@grasshopper.cs.duke.edu> <15735.47204.905352.900631@grasshopper.cs.duke.edu> <20020905201443.GD13050@cicely9.cicely.de> <15735.48281.30915.800894@grasshopper.cs.duke.edu> <20020905204558.GF13050@cicely9.cicely.de> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter writes: > On Thu, Sep 05, 2002 at 04:20:41PM -0400, Andrew Gallatin wrote: > > > > Bernd Walter writes: > > > OK - I have some basic understandig problems here. > > > > > > Why should ithreads ever return to PAL? > > > > Because PAL initiates the interrupt and creates a stackframe on the > > interrupted thread's kernel stack. It then calls the OSes interrupt > > vector. In order to restore the state of the world to be like it was > > before the interrupt happened, we need to return back to pal. > > That's clear, but this is not the ithread. > ithreads never return - see ithread_loop() in kern_intr.c. > I think we were being sloppy with our language. The code which schedules the ithread is what returns back to PAL. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:51:23 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 180C137B400; Thu, 5 Sep 2002 13:51:21 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4AA7243E3B; Thu, 5 Sep 2002 13:51:20 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA10174; Thu, 5 Sep 2002 16:51:17 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85Kolt58623; Thu, 5 Sep 2002 16:50:47 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.50087.139217.141521@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 16:50:47 -0400 (EDT) To: John Baldwin Cc: ticso@cicely.de, freebsd-alpha@FreeBSD.org Subject: Re: alpha performance on -current In-Reply-To: References: <15735.47813.266130.216534@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > OK, if the appended is what you had in mind, then I'm a much happier > > camper. > > Yep. > > > This patch: > > > > 8314.06 real 7025.32 user 956.14 sys > > > > Stock kernel: > > > > 8699.25 real 6985.64 user 1379.72 sys > > > > > > A 30% reduction in system time is pretty good. > > Cool. I'll be sure and try to test it on the DS20 I have to make > sure it is safe. Please do.. If its safe, please go ahead and commit it ;) Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 13:58:54 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4F17937B400 for ; Thu, 5 Sep 2002 13:58:52 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id E3A1843E3B for ; Thu, 5 Sep 2002 13:58:51 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 8827 invoked from network); 5 Sep 2002 20:58:49 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 5 Sep 2002 20:58:49 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g85KwnBv012536; Thu, 5 Sep 2002 16:58:49 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.50087.139217.141521@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 16:58:49 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: alpha performance on -current Cc: freebsd-alpha@FreeBSD.org, ticso@cicely.de Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > > > On 05-Sep-2002 Andrew Gallatin wrote: > > > > > > OK, if the appended is what you had in mind, then I'm a much happier > > > camper. > > > > Yep. > > > > > This patch: > > > > > > 8314.06 real 7025.32 user 956.14 sys > > > > > > Stock kernel: > > > > > > 8699.25 real 6985.64 user 1379.72 sys > > > > > > > > > A 30% reduction in system time is pretty good. > > > > Cool. I'll be sure and try to test it on the DS20 I have to make > > sure it is safe. > > Please do.. If its safe, please go ahead and commit it ;) Oh, go ahead and commit, it is current after all. :) I'll cvs up and test it tomorrow. :) -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 14: 8:43 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3338237B400; Thu, 5 Sep 2002 14:08:42 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 508A843E6A; Thu, 5 Sep 2002 14:08:41 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id RAA10805; Thu, 5 Sep 2002 17:08:40 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g85L8Ak58650; Thu, 5 Sep 2002 17:08:10 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15735.51130.562617.753358@grasshopper.cs.duke.edu> Date: Thu, 5 Sep 2002 17:08:10 -0400 (EDT) To: John Baldwin Cc: freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: References: <15735.50087.139217.141521@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin writes: > Oh, go ahead and commit, it is current after all. :) I'll cvs up and > test it tomorrow. :) I'd rather not commit code if it might be broken. I'd greatly prefer it if somebody with an MP box could try it.. I've left it in ~gallatin/atomic.patch on freefall, for easier patching. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 15:22:10 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 02D5037B400; Thu, 5 Sep 2002 15:22:08 -0700 (PDT) Received: from canning.wemm.org (canning.wemm.org [192.203.228.65]) by mx1.FreeBSD.org (Postfix) with ESMTP id C184F43E4A; Thu, 5 Sep 2002 15:22:07 -0700 (PDT) (envelope-from peter@wemm.org) Received: from wemm.org (localhost [127.0.0.1]) by canning.wemm.org (Postfix) with ESMTP id A8AB82A88D; Thu, 5 Sep 2002 15:22:07 -0700 (PDT) (envelope-from peter@wemm.org) X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4 To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: alpha performance on -current In-Reply-To: <15735.51130.562617.753358@grasshopper.cs.duke.edu> Date: Thu, 05 Sep 2002 15:22:07 -0700 From: Peter Wemm Message-Id: <20020905222207.A8AB82A88D@canning.wemm.org> Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Andrew Gallatin wrote: > > John Baldwin writes: > > Oh, go ahead and commit, it is current after all. :) I'll cvs up and > > test it tomorrow. :) > > I'd rather not commit code if it might be broken. I'd greatly prefer > it if somebody with an MP box could try it.. > > I've left it in ~gallatin/atomic.patch on freefall, for easier > patching. I can give it a spin on an SMP UP2000 shortly. Cheers, -Peter -- Peter Wemm - peter@wemm.org; peter@FreeBSD.org; peter@yahoo-inc.com "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 18:38:33 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6EF0E37B400; Thu, 5 Sep 2002 18:38:31 -0700 (PDT) Received: from avocet.mail.pas.earthlink.net (avocet.mail.pas.earthlink.net [207.217.120.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1767643E65; Thu, 5 Sep 2002 18:38:31 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0142.cvx22-bradley.dialup.earthlink.net ([209.179.198.142] helo=mindspring.com) by avocet.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17n84i-0005rE-00; Thu, 05 Sep 2002 18:38:21 -0700 Message-ID: <3D7806D1.C87A9A6A@mindspring.com> Date: Thu, 05 Sep 2002 18:37:21 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Andrew Gallatin Cc: John Baldwin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption References: <15735.47204.905352.900631@grasshopper.cs.duke.edu> <15735.47756.501169.199225@grasshopper.cs.duke.edu> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Andrew Gallatin wrote: > John Baldwin writes: > > Solaris doesn't run on alpha, but it also a bit different in its approach. > > I do wonder if there is a way we can violate an assumption in PAL due to > > migration though. That is, a thread could return to PAL on a different > > CPU than the one the interrupt was originally sent to. This might explain > > why only SMP has problems. > > > > Hey ... I think you have it on the nose! That makes the most sense > I've heard yet. > > Do we have any way to bind a thread to a cpu? Alfred has some patches. IMO, threads should not be so bound to a CPU, as a migration event should be an exception, rather than the rule, and always handled by the scheduler. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 18:50: 6 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8778537B400; Thu, 5 Sep 2002 18:50:00 -0700 (PDT) Received: from avocet.mail.pas.earthlink.net (avocet.mail.pas.earthlink.net [207.217.120.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1D30143E4A; Thu, 5 Sep 2002 18:50:00 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0142.cvx22-bradley.dialup.earthlink.net ([209.179.198.142] helo=mindspring.com) by avocet.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17n8Fz-0006Do-00; Thu, 05 Sep 2002 18:49:59 -0700 Message-ID: <3D78098B.CEBF13EC@mindspring.com> Date: Thu, 05 Sep 2002 18:48:59 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: John Baldwin Cc: Andrew Gallatin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org John Baldwin wrote: > > Hey ... I think you have it on the nose! That makes the most sense > > I've heard yet. > > > > Do we have any way to bind a thread to a cpu? > > I used to. Then KSE3 was committed. I suppose I could rewrite it from > scratch again. The ithreads are different from threads. People don't seem to get this idea, but it is nevertheless true. The x86 SMP doesn't have this problem (yet), in that the IRQ is disabled on the I/O APIC, so it doesn't matter what CPU it runs on, it's a seperate thing. If x86 NUMA systems are ever supported, or in some cases the AMD Alpha-derived arbitration is used, there could be a problem (but not for the same reasons as on the Alpha). Using all the same primitives for ithreads as for threads will lead to things like a shared IRQ which comes true for multiple devices will need to count down to enable, if the ithreads for servicing are run simultaneously on seperate CPUs. This also implies a global lock on enabling a newly arrived device that's sharing an already allocated IRQ, and inter-CPU contnetion on any IRQ whose use cont is != 1. Interrupt threads make some very messy thing elegant, but they do so at the cost of moving the mess elsewhere. If you have to characterize it, call it "The law of conservation of inelegance". 8-). The only thing that ever saves you from it is a top-down design. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 21: 7:31 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3D20637B400; Thu, 5 Sep 2002 21:07:27 -0700 (PDT) Received: from obsecurity.dyndns.org (adsl-64-165-226-88.dsl.lsan03.pacbell.net [64.165.226.88]) by mx1.FreeBSD.org (Postfix) with ESMTP id 40E1543E7B; Thu, 5 Sep 2002 21:06:47 -0700 (PDT) (envelope-from kris@obsecurity.org) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 99DE266D7A; Thu, 5 Sep 2002 21:06:41 -0700 (PDT) Date: Thu, 5 Sep 2002 21:06:41 -0700 From: Kris Kennaway To: gnome@FreeBSD.org, alpha@FreeBSD.org Cc: re@FreeBSD.org Subject: mozilla failure on 4.6.2 Message-ID: <20020906040641.GA61111@xor.obsecurity.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="PNTmBPCT7hxwcZjr" Content-Disposition: inline User-Agent: Mutt/1.4i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org --PNTmBPCT7hxwcZjr Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Anyone able to take a look? http://beta.freebsd.org/errorlogs/4-latest/mozilla-1.0.rc3,1.log Kris --PNTmBPCT7hxwcZjr Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (FreeBSD) iD8DBQE9eCnQWry0BWjoQKURAi41AJ4nmigG+WWgjPw8cw03h1vYAo/tPgCguooW eOVK3i+0GsOSWYcIydh2YX4= =+WZs -----END PGP SIGNATURE----- --PNTmBPCT7hxwcZjr-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 21:18:37 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B335337B400; Thu, 5 Sep 2002 21:18:31 -0700 (PDT) Received: from obsecurity.dyndns.org (adsl-64-165-226-88.dsl.lsan03.pacbell.net [64.165.226.88]) by mx1.FreeBSD.org (Postfix) with ESMTP id AF8CF43EC5; Thu, 5 Sep 2002 21:16:23 -0700 (PDT) (envelope-from kris@obsecurity.org) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 2CC5C66D7A; Thu, 5 Sep 2002 21:16:13 -0700 (PDT) Date: Thu, 5 Sep 2002 21:16:12 -0700 From: Kris Kennaway To: re@FreeBSD.org, kde@FreeBSD.org Cc: alpha@FreeBSD.org Subject: kde build broken on alpha 4.6.2 Message-ID: <20020906041612.GA61205@xor.obsecurity.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="qDbXVdCdHGoSgWSk" Content-Disposition: inline User-Agent: Mutt/1.4i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org --qDbXVdCdHGoSgWSk Content-Type: text/plain; charset=us-ascii Content-Disposition: inline See: http://beta.freebsd.org/errorlogs/4-latest/kdeadmin-3.0_2.log http://beta.freebsd.org/errorlogs/4-latest/kdegraphics-3.0_2.log These are necessary for making KDE available for the alpha 4.6.2-RELEASE packages. Will make it abundantly clear that he doesn't care about fixing this; perhaps someone else can take a look. Kris --qDbXVdCdHGoSgWSk Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.7 (FreeBSD) iD8DBQE9eCwKWry0BWjoQKURAqHRAJ9O0CEW61917fjBWcjZMx7QBJ2F2gCglU4Z NWOv6lvNE/lgi5G7J2M5K8s= =Zoc3 -----END PGP SIGNATURE----- --qDbXVdCdHGoSgWSk-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 21:34:39 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6958337B401; Thu, 5 Sep 2002 21:34:35 -0700 (PDT) Received: from procyon.firepipe.net (procyon.firepipe.net [198.78.66.151]) by mx1.FreeBSD.org (Postfix) with ESMTP id 112F643E4A; Thu, 5 Sep 2002 21:34:35 -0700 (PDT) (envelope-from will@csociety.org) Received: by procyon.firepipe.net (Postfix, from userid 1000) id AE05B2145B; Thu, 5 Sep 2002 21:32:51 -0700 (PDT) Date: Thu, 5 Sep 2002 21:32:51 -0700 From: Will Andrews To: Kris Kennaway Cc: re@FreeBSD.org, kde@FreeBSD.org, alpha@FreeBSD.org Subject: Re: [kde-freebsd] kde build broken on alpha 4.6.2 Message-ID: <20020906043251.GV96003@procyon.firepipe.net> Mail-Followup-To: Will Andrews , Kris Kennaway , re@FreeBSD.org, kde@FreeBSD.org, alpha@FreeBSD.org References: <20020906041612.GA61205@xor.obsecurity.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20020906041612.GA61205@xor.obsecurity.org> User-Agent: Mutt/1.4i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 09:16:12PM -0700, Kris Kennaway wrote: > http://beta.freebsd.org/errorlogs/4-latest/kdeadmin-3.0_2.log > http://beta.freebsd.org/errorlogs/4-latest/kdegraphics-3.0_2.log > > These are necessary for making KDE available for the alpha > 4.6.2-RELEASE packages. > > Will make it abundantly clear that he doesn't care about fixing this; > perhaps someone else can take a look. That is not correct. I said it's not worth my time to fix it. You must remember that "worth" in this case includes needing to know why the compiler is misbehaving, and needing a sufficiently properly set up tree on an alpha box. Then, and only then, might I actually get some progress on this problem. Not to mention that the packages shipping in this version are way out of date and people can still use KDE without the above packages. Please don't ever publically misrepresent me again. Thanks. I should note that I'm tired of being blamed for alpha-specific problems when the alpha people should be the ones fixing it, if they so desire. AFAIK, there is nobody on the kde@ list who has an alpha so that platform continues to enjoy relatively poor support for KDE on FreeBSD. Regards, -- wca To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 21:46:57 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id F28DB37B400; Thu, 5 Sep 2002 21:46:53 -0700 (PDT) Received: from creme-brulee.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2302243E65; Thu, 5 Sep 2002 21:46:53 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from gyros (gyros.marcuscom.com [192.168.1.9]) by creme-brulee.marcuscom.com (8.12.5/8.12.5) with ESMTP id g864j2ix073546; Fri, 6 Sep 2002 00:45:02 -0400 (EDT) (envelope-from marcus@marcuscom.com) Subject: Re: mozilla failure on 4.6.2 From: Joe Marcus Clarke To: Kris Kennaway Cc: gnome@FreeBSD.ORG, alpha@FreeBSD.ORG, re@FreeBSD.ORG In-Reply-To: <20020906040641.GA61111@xor.obsecurity.org> References: <20020906040641.GA61111@xor.obsecurity.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 00:46:38 -0400 Message-Id: <1031287598.361.2.camel@gyros.marcuscom.com> Mime-Version: 1.0 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 2002-09-06 at 00:06, Kris Kennaway wrote: > Anyone able to take a look? > > http://beta.freebsd.org/errorlogs/4-latest/mozilla-1.0.rc3,1.log This is "fixed" now. The reason I say "fixed" is that Mozilla is currently broken on Alpha FreeBSD. The problem appears to be issues with Alpha assembly code in our version of the GCC compiler. However, your immediate problem can be solved by removing the CONFIGURE_ARG --enable-reorder from the mozilla Makefile. Joe > > Kris -- PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Thu Sep 5 23: 6:50 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 740A937B400; Thu, 5 Sep 2002 23:06:47 -0700 (PDT) Received: from smtp.web.de (smtp03.web.de [217.72.192.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9614E43E3B; Thu, 5 Sep 2002 23:06:46 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [217.82.34.37] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17nCGT-00069z-00; Fri, 06 Sep 2002 08:06:45 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 6EA7D29C; Fri, 6 Sep 2002 08:06:41 +0200 (CEST) Received: from jan-freebsd.lan (jan-freebsd.lan [192.168.0.22]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id C7AE928F; Fri, 6 Sep 2002 08:06:10 +0200 (CEST) Subject: Re: kde build broken on alpha 4.6.2 From: Jan Lentfer To: Kris Kennaway Cc: re@FreeBSD.org, kde@FreeBSD.org, alpha@FreeBSD.org In-Reply-To: <20020906041612.GA61205@xor.obsecurity.org> References: <20020906041612.GA61205@xor.obsecurity.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 08:06:09 +0200 Message-Id: <1031292369.244.4.camel@jan-freebsd.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Am Fr, 2002-09-06 um 06.16 schrieb Kris Kennaway: > See: > > http://beta.freebsd.org/errorlogs/4-latest/kdeadmin-3.0_2.log > http://beta.freebsd.org/errorlogs/4-latest/kdegraphics-3.0_2.log > > These are necessary for making KDE available for the alpha > 4.6.2-RELEASE packages. > > Will make it abundantly clear that he doesn't care about fixing this; > perhaps someone else can take a look. I build kde 3.0.3 from ports on a PWS 500a (21164A CPU) 2 or three days ago after a cvsup of the ports tree - I didn't encounter any errors. kdegraphics is there, kdeadmin not. I'll try to build kdeadmin seperatly now. bash-2.05a# pkg_info | grep kde kde-3.0.3 The "meta-port" for KDE kdebase-3.0.3 Base modules for the KDE integrated X11 desktop kdegames-3.0.3 Games for the KDE integrated X11 desktop kdegraphics-3.0.3 Graphics utilities for the KDE3 integrated X11 desktop kdelibs-3.0.3 Libraries for KDE kdemultimedia-3.0.3 Multimedia utilities for the KDE integrated X11 desktop kdenetwork-3.0.3 Network modules for KDE3 kdeutils-3.0.3 Utilities for the KDE integrated X11 desktop Jan -- Jan Lentfer System Administrator Molecular Cell Biology / AG Holstein, Darmstadt University of Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 4712037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 0: 2: 8 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0EF7037B400 for ; Fri, 6 Sep 2002 00:02:06 -0700 (PDT) Received: from freebie.xs4all.nl (freebie.xs4all.nl [213.84.32.253]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5CF8243E75 for ; Fri, 6 Sep 2002 00:02:04 -0700 (PDT) (envelope-from wkb@freebie.xs4all.nl) Received: from freebie.xs4all.nl (localhost [127.0.0.1]) by freebie.xs4all.nl (8.12.5/8.12.2) with ESMTP id g86723H7009740 for ; Fri, 6 Sep 2002 09:02:03 +0200 (CEST) (envelope-from wkb@freebie.xs4all.nl) Received: (from wkb@localhost) by freebie.xs4all.nl (8.12.5/8.12.5/Submit) id g86723Do009739 for freebsd-alpha@freebsd.org; Fri, 6 Sep 2002 09:02:03 +0200 (CEST) Date: Fri, 6 Sep 2002 09:02:03 +0200 From: Wilko Bulte To: freebsd-alpha@freebsd.org Subject: 4.6.2R mini-iso up for ftp now Message-ID: <20020906090203.A9723@freebie.xs4all.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i X-OS: FreeBSD 4.6-STABLE X-PGP: finger wilko@freebsd.org Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org See ftp.freebsd.org:/pub/FreeBSD/releases/alpha/ISO-IMAGES/4.6.2/4.6.2-mini.iso -rw-r--r-- 1 ftpadmin ftpadmin 196870144 3 Sep 20:29 4.6.2-mini.iso -- | / o / /_ _ wilko@FreeBSD.org |/|/ / / /( (_) Bulte Arnhem, the Netherlands To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 1:43:15 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0DD0737B400; Fri, 6 Sep 2002 01:43:13 -0700 (PDT) Received: from smtp.web.de (smtp01.web.de [194.45.170.210]) by mx1.FreeBSD.org (Postfix) with ESMTP id DD65343E72; Fri, 6 Sep 2002 01:43:11 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [217.82.34.37] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17nDn1-0004rz-00; Fri, 06 Sep 2002 09:44:27 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 1F91129C; Fri, 6 Sep 2002 09:44:23 +0200 (CEST) Received: from jan-freebsd.lan (jan-freebsd.lan [192.168.0.22]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id A50FE28F; Fri, 6 Sep 2002 09:44:02 +0200 (CEST) Subject: Re: kde build broken on alpha 4.6.2 From: Jan Lentfer Cc: Kris Kennaway , re@FreeBSD.org, kde@FreeBSD.org, alpha@FreeBSD.org In-Reply-To: <1031292369.244.4.camel@jan-freebsd.lan> References: <20020906041612.GA61205@xor.obsecurity.org> <1031292369.244.4.camel@jan-freebsd.lan> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 09:44:00 +0200 Message-Id: <1031298241.244.7.camel@jan-freebsd.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Am Fr, 2002-09-06 um 08.06 schrieb Jan Lentfer: > > I build kde 3.0.3 from ports on a PWS 500a (21164A CPU) 2 or three days > ago after a cvsup of the ports tree - I didn't encounter any errors. > kdegraphics is there, kdeadmin not. I'll try to build kdeadmin seperatly > now. kdeadmin build and installed without problems, too. I don't know why it wasn't installed along with the metaport... Jan -- Jan Lentfer System Administrator Molecular Cell Biology / AG Holstein, Darmstadt University of Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 4712037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 1:55:26 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8840637B400 for ; Fri, 6 Sep 2002 01:55:23 -0700 (PDT) Received: from crucible.athame.co.uk (guru164.netsonic.fi [194.29.193.164]) by mx1.FreeBSD.org (Postfix) with ESMTP id E3CE543E3B for ; Fri, 6 Sep 2002 01:55:22 -0700 (PDT) (envelope-from andy@athame.co.uk) Received: from amavis by crucible.athame.co.uk with scanned-ok (Exim 3.33 #1) id 17nEtd-00006u-00 for alpha@FreeBSD.org; Fri, 06 Sep 2002 11:55:21 +0300 Received: from zappa.int.athame.co.uk ([192.168.1.3] helo=FSOPTI208.athame.co.uk) by crucible.athame.co.uk with esmtp (Exim 3.33 #1) id 17nEsr-00005w-00; Fri, 06 Sep 2002 11:54:34 +0300 Message-Id: <5.1.0.14.0.20020906114849.028f3070@localhost> X-Sender: tap@localhost X-Mailer: QUALCOMM Windows Eudora Version 5.1 Date: Fri, 06 Sep 2002 11:54:33 +0300 To: Jan Lentfer From: Andy Fawcett Subject: Re: [kde-freebsd] Re: kde build broken on alpha 4.6.2 Cc: Kris Kennaway , re@FreeBSD.org, kde@FreeBSD.org, alpha@FreeBSD.org In-Reply-To: <1031298241.244.7.camel@jan-freebsd.lan> References: <1031292369.244.4.camel@jan-freebsd.lan> <20020906041612.GA61205@xor.obsecurity.org> <1031292369.244.4.camel@jan-freebsd.lan> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed X-Virus-Scanned: by AMaViS perl-11 at crucible.athame.co.uk Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org At 10:44 06/09/2002, Jan Lentfer wrote: >Am Fr, 2002-09-06 um 08.06 schrieb Jan Lentfer: > > > > > I build kde 3.0.3 from ports on a PWS 500a (21164A CPU) 2 or three days > > ago after a cvsup of the ports tree - I didn't encounter any errors. > > kdegraphics is there, kdeadmin not. I'll try to build kdeadmin seperatly > > now. > >kdeadmin build and installed without problems, too. I don't know why it >wasn't installed along with the metaport... Thanks for the feedback, it's good to know that 3.0.3 works on Alpha. As for the reason why the meta-port doesn't install kdeadmin3, it's because we made a conscious decision not to blindly install all of KDE from it. At the moment, we install kdegames, kdeutils, koffice, kdenetwork, kdegraphics, kdemultimedia in addition to the minimum requirements of arts, kdelibs, and kdebase. This leaves kdeadmin, kdeaddons, kdeartwork, kdeedu, kdepim, kdesdk, and kdetoys uninstalled. The choices were made because these last few are the ones least likely to be needed by everyone, and space on the RELEASE CDs is at a premium. Of course, YMMV :) Regards, Andy (of the KDE-FreeBSD team, but not speaking for them) -- Andy Fawcett andy@athame.co.uk To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 2: 5:58 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 026DB37B400; Fri, 6 Sep 2002 02:05:56 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id B707343E4A; Fri, 6 Sep 2002 02:05:54 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g8695l6K081217 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Fri, 6 Sep 2002 11:05:49 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g8695kNk005614 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Fri, 6 Sep 2002 11:05:46 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g8695JD3015015; Fri, 6 Sep 2002 11:05:19 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g8695IHk015014; Fri, 6 Sep 2002 11:05:18 +0200 (CEST) Date: Fri, 6 Sep 2002 11:05:18 +0200 From: Bernd Walter To: Terry Lambert Cc: John Baldwin , Andrew Gallatin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption Message-ID: <20020906090517.GI13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <3D78098B.CEBF13EC@mindspring.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3D78098B.CEBF13EC@mindspring.com> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, Sep 05, 2002 at 06:48:59PM -0700, Terry Lambert wrote: > John Baldwin wrote: > > > Hey ... I think you have it on the nose! That makes the most sense > > > I've heard yet. > > > > > > Do we have any way to bind a thread to a cpu? > > > > I used to. Then KSE3 was committed. I suppose I could rewrite it from > > scratch again. > > The ithreads are different from threads. People don't seem to > get this idea, but it is nevertheless true. > > The x86 SMP doesn't have this problem (yet), in that the IRQ > is disabled on the I/O APIC, so it doesn't matter what CPU it > runs on, it's a seperate thing. Interrupts are disabled globaly on alpha too. The only platform where we disable on the CPU is the PC164 as a workaround, but this system is UP. > If x86 NUMA systems are ever supported, or in some cases the > AMD Alpha-derived arbitration is used, there could be a problem > (but not for the same reasons as on the Alpha). I expect ithreads to be one of the less critical points on NUMA. > Using all the same primitives for ithreads as for threads will > lead to things like a shared IRQ which comes true for multiple > devices will need to count down to enable, if the ithreads for > servicing are run simultaneously on seperate CPUs. This also > implies a global lock on enabling a newly arrived device that's > sharing an already allocated IRQ, and inter-CPU contnetion on > any IRQ whose use cont is != 1. Currently shared interrupts also share an ithread. > Interrupt threads make some very messy thing elegant, but they > do so at the cost of moving the mess elsewhere. If you have to > characterize it, call it "The law of conservation of inelegance". > 8-). The only thing that ever saves you from it is a top-down > design. Possible, at least that's our current design. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 4:33:57 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 956B037B400; Fri, 6 Sep 2002 04:33:52 -0700 (PDT) Received: from swan.mail.pas.earthlink.net (swan.mail.pas.earthlink.net [207.217.120.123]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2F4EF43E6A; Fri, 6 Sep 2002 04:33:52 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0015.cvx21-bradley.dialup.earthlink.net ([209.179.192.15] helo=mindspring.com) by swan.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17nHMv-0002Ui-00; Fri, 06 Sep 2002 04:33:45 -0700 Message-ID: <3D78925A.DAA13463@mindspring.com> Date: Fri, 06 Sep 2002 04:32:42 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: ticso@cicely.de Cc: John Baldwin , Andrew Gallatin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption References: <3D78098B.CEBF13EC@mindspring.com> <20020906090517.GI13050@cicely9.cicely.de> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter wrote: > Interrupts are disabled globaly on alpha too. > The only platform where we disable on the CPU is the PC164 as > a workaround, but this system is UP. This misses the point. The point is whether or not ithreads run to completion on the processor to which the interrupt is delivered. The problem in this case on the Alpha is that interrupts are routed through the PAL code on a particular processor, and so the return has to be to the PAL code on the same processor, because there is a context cons'ed up for it that have to be destructed on the same CPU where it was cons'ed. > > If x86 NUMA systems are ever supported, or in some cases the > > AMD Alpha-derived arbitration is used, there could be a problem > > (but not for the same reasons as on the Alpha). > > I expect ithreads to be one of the less critical points on NUMA. The problem in this case is that you have to do completion counting on the ISR's for any given interrupt. For example, on an SMP box running two CPUs where IRQ A is a shared interrupt for two devices, then if you want to dispatch for one device to the first CPU, and the second device to the second, then when you reenable interrupts depends on all ISRs having run to completion. So you have to set a global count to "2", for the number of ISRs that have to run, and then run each one on a CPU, decrement the count, and when the count goes from 1->0, reenable the interrupt. For NUMA, it really depends on the cluster architecture. If you have devices associated based on CPU clusters, that's one thing; it's an easy call. If you have it on the basis of adjacency, it is not so easy a call, because the adjacency det for two different devices can be only partially intersecting (i.e. dev 1 is associated with CPU's 1,2,3,4, dev 2 with CPU's 5,6,7,8, and dev 3 with CPU's 3,4,5,6). This gets into the same issue as the Alpha, where the CPU to take the interrupt has to complete the interrupt, only in this case, you are talking abbout the associativity set. Yeah, most likely this won't be a problem, but then that's likely the same thing that was thought when the current Alpha problem was introduced. > > Using all the same primitives for ithreads as for threads will > > lead to things like a shared IRQ which comes true for multiple > > devices will need to count down to enable, if the ithreads for > > servicing are run simultaneously on seperate CPUs. This also > > implies a global lock on enabling a newly arrived device that's > > sharing an already allocated IRQ, and inter-CPU contnetion on > > any IRQ whose use cont is != 1. > > Currently shared interrupts also share an ithread. Yeah; this isn't very efficient with only 4 interrupts and a lot of PCI cards. > > Interrupt threads make some very messy things elegant, but they > > do so at the cost of moving the mess elsewhere. If you have to > > characterize it, call it "The law of conservation of inelegance". > > 8-). The only thing that ever saves you from it is a top-down > > design. > > Possible, at least that's our current design. I'm not sure if this is true. As you said, currently, shared interrupts also share an ithread. Thread affinity as an explicit hard0coded attribute is probably not the correct fix for the current Alpha probems. It will make it harder to do it right later (just like it's harder to fix a foundation after you've built a house on it). -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 5: 1: 4 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DC4E637B400; Fri, 6 Sep 2002 05:00:58 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id CA2F443E75; Fri, 6 Sep 2002 05:00:57 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g86C0i6K082274 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Fri, 6 Sep 2002 14:00:48 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g86C0gNk006366 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Fri, 6 Sep 2002 14:00:43 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g86C0ED3017145; Fri, 6 Sep 2002 14:00:14 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g86C0CH4017144; Fri, 6 Sep 2002 14:00:12 +0200 (CEST) Date: Fri, 6 Sep 2002 14:00:12 +0200 From: Bernd Walter To: Terry Lambert Cc: ticso@cicely.de, John Baldwin , Andrew Gallatin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption Message-ID: <20020906120011.GO13050@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <3D78098B.CEBF13EC@mindspring.com> <20020906090517.GI13050@cicely9.cicely.de> <3D78925A.DAA13463@mindspring.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3D78925A.DAA13463@mindspring.com> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, Sep 06, 2002 at 04:32:42AM -0700, Terry Lambert wrote: > Bernd Walter wrote: > > Interrupts are disabled globaly on alpha too. > > The only platform where we disable on the CPU is the PC164 as > > a workaround, but this system is UP. > > This misses the point. The point is whether or not ithreads > run to completion on the processor to which the interrupt is > delivered. What I missed is why you think this is different with APICs on i386. Well I have to say that I don't know much about APICs. > The problem in this case on the Alpha is that interrupts are > routed through the PAL code on a particular processor, and > so the return has to be to the PAL code on the same processor, > because there is a context cons'ed up for it that have to be > destructed on the same CPU where it was cons'ed. That's the most logical (if not the only) theorie so far. > > > If x86 NUMA systems are ever supported, or in some cases the > > > AMD Alpha-derived arbitration is used, there could be a problem > > > (but not for the same reasons as on the Alpha). > > > > I expect ithreads to be one of the less critical points on NUMA. > > The problem in this case is that you have to do completion > counting on the ISR's for any given interrupt. For example, > on an SMP box running two CPUs where IRQ A is a shared interrupt > for two devices, then if you want to dispatch for one device to > the first CPU, and the second device to the second, then when > you reenable interrupts depends on all ISRs having run to > completion. So you have to set a global count to "2", for the > number of ISRs that have to run, and then run each one on a CPU, > decrement the count, and when the count goes from 1->0, reenable > the interrupt. If you want more than one ithread per intline I agree. > For NUMA, it really depends on the cluster architecture. If you > have devices associated based on CPU clusters, that's one thing; > it's an easy call. If you have it on the basis of adjacency, it > is not so easy a call, because the adjacency det for two different > devices can be only partially intersecting (i.e. dev 1 is associated > with CPU's 1,2,3,4, dev 2 with CPU's 5,6,7,8, and dev 3 with CPU's > 3,4,5,6). This gets into the same issue as the Alpha, where the > CPU to take the interrupt has to complete the interrupt, only in > this case, you are talking abbout the associativity set. I see. > Yeah, most likely this won't be a problem, but then that's likely > the same thing that was thought when the current Alpha problem > was introduced. Ack. > > > Using all the same primitives for ithreads as for threads will > > > lead to things like a shared IRQ which comes true for multiple > > > devices will need to count down to enable, if the ithreads for > > > servicing are run simultaneously on seperate CPUs. This also > > > implies a global lock on enabling a newly arrived device that's > > > sharing an already allocated IRQ, and inter-CPU contnetion on > > > any IRQ whose use cont is != 1. > > > > Currently shared interrupts also share an ithread. > > Yeah; this isn't very efficient with only 4 interrupts and a lot > of PCI cards. After all you are writing to the alpha list :) On alphas we typically have 4 intlines per slot on the primary busses. Only small machines like LCA share 4 lines for all slots. As typical chips take only one intline we even don't share intlines over PCI-PCI bridges with up to 4 chips. But generally I see the point that when the handling for one device is blocked the service for devices sharing the same intline are also blocked - there is a good reason that blocking in device drivers has to be short timed. > > > Interrupt threads make some very messy things elegant, but they > > > do so at the cost of moving the mess elsewhere. If you have to > > > characterize it, call it "The law of conservation of inelegance". > > > 8-). The only thing that ever saves you from it is a top-down > > > design. > > > > Possible, at least that's our current design. > > I'm not sure if this is true. As you said, currently, shared > interrupts also share an ithread. > > Thread affinity as an explicit hard0coded attribute is probably > not the correct fix for the current Alpha probems. It will make > it harder to do it right later (just like it's harder to fix a > foundation after you've built a house on it). I can't ague on that yet, because I understand the reason why returning back to PAL could be on a different CPU only since a few hours. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 5:28: 3 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D411937B400 for ; Fri, 6 Sep 2002 05:28:01 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0D55F43E42 for ; Fri, 6 Sep 2002 05:27:57 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id IAA01749; Fri, 6 Sep 2002 08:27:56 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86CRQ459828; Fri, 6 Sep 2002 08:27:26 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15736.40749.986834.818558@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 08:27:25 -0400 (EDT) To: Joe Marcus Clarke Cc: alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <1031287598.361.2.camel@gyros.marcuscom.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Joe Marcus Clarke writes: > On Fri, 2002-09-06 at 00:06, Kris Kennaway wrote: > > Anyone able to take a look? > > > > http://beta.freebsd.org/errorlogs/4-latest/mozilla-1.0.rc3,1.log > > This is "fixed" now. The reason I say "fixed" is that Mozilla is > currently broken on Alpha FreeBSD. The problem appears to be issues > with Alpha assembly code in our version of the GCC compiler. Can you elaborate, please? Thanks, Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 5:37:59 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 08E4D37B400; Fri, 6 Sep 2002 05:37:53 -0700 (PDT) Received: from harrier.mail.pas.earthlink.net (harrier.mail.pas.earthlink.net [207.217.120.12]) by mx1.FreeBSD.org (Postfix) with ESMTP id 454D443E6A; Fri, 6 Sep 2002 05:37:52 -0700 (PDT) (envelope-from tlambert2@mindspring.com) Received: from pool0015.cvx21-bradley.dialup.earthlink.net ([209.179.192.15] helo=mindspring.com) by harrier.mail.pas.earthlink.net with esmtp (Exim 3.33 #1) id 17nIMl-00052K-00; Fri, 06 Sep 2002 05:37:39 -0700 Message-ID: <3D78A148.F25A8F27@mindspring.com> Date: Fri, 06 Sep 2002 05:36:24 -0700 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: ticso@cicely.de Cc: John Baldwin , Andrew Gallatin , freebsd-alpha@FreeBSD.ORG Subject: Re: ithread preemption References: <3D78098B.CEBF13EC@mindspring.com> <20020906090517.GI13050@cicely9.cicely.de> <3D78925A.DAA13463@mindspring.com> <20020906120011.GO13050@cicely9.cicely.de> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Bernd Walter wrote: > > > Interrupts are disabled globaly on alpha too. > > > The only platform where we disable on the CPU is the PC164 as > > > a workaround, but this system is UP. > > > > This misses the point. The point is whether or not ithreads > > run to completion on the processor to which the interrupt is > > delivered. > > What I missed is why you think this is different with APICs on i386. > Well I have to say that I don't know much about APICs. Ah. Because even though Intel recommends SMP systems run in virtual wire mode, FreeBSD does not run in virtual wire mode. Instead, grabbing the giant grabs the interrupt (in older SMP code). In the SMPng code, the interrupt routing is explicitly managed, and it's still not an issue. > > The problem in this case on the Alpha is that interrupts are > > routed through the PAL code on a particular processor, and > > so the return has to be to the PAL code on the same processor, > > because there is a context cons'ed up for it that have to be > > destructed on the same CPU where it was cons'ed. > > That's the most logical (if not the only) theorie so far. It's the only one I've seen. Even if it turns out not to be the cause of the particular problem, it *could* cause problems, according to the PAL documentation (what there is of it) online. > > > I expect ithreads to be one of the less critical points on NUMA. > > > > The problem in this case is that you have to do completion > > counting on the ISR's for any given interrupt. For example, > > on an SMP box running two CPUs where IRQ A is a shared interrupt > > for two devices, then if you want to dispatch for one device to > > the first CPU, and the second device to the second, then when > > you reenable interrupts depends on all ISRs having run to > > completion. So you have to set a global count to "2", for the > > number of ISRs that have to run, and then run each one on a CPU, > > decrement the count, and when the count goes from 1->0, reenable > > the interrupt. > > If you want more than one ithread per intline I agree. I think it may be a requirement in the future. I don't think spreading them out across intlines is going to work, unless you can either wire them down (the PCI code allows it, I guess), and/or have some config way of specifying an assignment preference (e.g. "Whatever you do, don't share an IRQ between the two Gigabit Etherenet cards", or "Whatever you do, don't wshare an interrupt between the ethernet card and the disck controller", etc.). I guess there is always card shuffling. 8-(. > > For NUMA, it really depends on the cluster architecture. If you > > have devices associated based on CPU clusters, that's one thing; > > it's an easy call. If you have it on the basis of adjacency, it > > is not so easy a call, because the adjacency det for two different > > devices can be only partially intersecting (i.e. dev 1 is associated > > with CPU's 1,2,3,4, dev 2 with CPU's 5,6,7,8, and dev 3 with CPU's > > 3,4,5,6). This gets into the same issue as the Alpha, where the > > CPU to take the interrupt has to complete the interrupt, only in > > this case, you are talking abbout the associativity set. > > I see. I are a geek. 8-) 8-). Actually, I have a real interest in seeing FreeBSD make it onto NUMA hardware, because (IMO), it's a hop, skip, and a jump to distributed processing, and I think the same problems will need solving. Of course, not everyone agrees with me, so assign a weighting factor below 1.0 to my opinion on what's important in this respect. > > Yeah, most likely this won't be a problem, but then that's likely > > the same thing that was thought when the current Alpha problem > > was introduced. > > Ack. I didn't mean it was on purpose, just that it was probably not something that someone really spent time thnking about before changing the code. > > > Currently shared interrupts also share an ithread. > > > > Yeah; this isn't very efficient with only 4 interrupts and a lot > > of PCI cards. > > After all you are writing to the alpha list :) > On alphas we typically have 4 intlines per slot on the primary busses. > Only small machines like LCA share 4 lines for all slots. > As typical chips take only one intline we even don't share intlines > over PCI-PCI bridges with up to 4 chips. > But generally I see the point that when the handling for one device > is blocked the service for devices sharing the same intline are also > blocked - there is a good reason that blocking in device drivers has > to be short timed. I think the code is going to end up shared, even if you are running the good DEC PCI chipsets, instead of the less able Intel ones. What that means to me is that when there is performance pressure, this is the type of change that will be made for the Intel side, and the Alpha will quit working (again). > > Thread affinity as an explicit hard-coded attribute is probably > > not the correct fix for the current Alpha probems. It will make > > it harder to do it right later (just like it's harder to fix a > > foundation after you've built a house on it). > > I can't ague on that yet, because I understand the reason why returning > back to PAL could be on a different CPU only since a few hours. My problem with hard-coding is that it will leave artifacts; my own answer to this would be to set a "don't migrate" flag, rather than a "run only on CPU X" flag. This will work if you have some scheduler cooperation, and will fall out naturally without having to change a lot of code. It requires per CPU run queues, though (gives you natural affinity anyway, where migration has to be done explicitly if it's to happen). The plus side is that you can get rid of all the global scheduler locks, and even on migration, if you push processes, rather than pull them, you can check for an empty push queue without a lock, and any locking you do will end up giving you at most a 2 CPU contention domain, instead of an N CPU contention (minor details like a "figure of merit boost" while in the scheduler, etc. can be handled later). If you wanted an initial "don't run on other CPU" flag, you could get to 32 CPU's pretty fast with a "run on" bitmap, and this would not be painful to migrate, like changing a lot of code would be, as long as you had a CPU ID to use as a shift index (I would use a 32 element array to get the bit value out with one add and compare instead of a shift, but that's probably premature optimization. 8-)). Basically, you add an "int" to the proc struct, and set the bit for the CPU you want to run on, and the current scheduler leaves your process at the head and skips over it. Inefficient, but effictive, for a proof-of-concept. I don't have a multiple CPU system that will run -current, or I'd send a patch. 8-(. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 8: 2:45 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 5ADCD37B400 for ; Fri, 6 Sep 2002 08:02:42 -0700 (PDT) Received: from mail.speakeasy.net (mail15.speakeasy.net [216.254.0.215]) by mx1.FreeBSD.org (Postfix) with ESMTP id F027C43E3B for ; Fri, 6 Sep 2002 08:02:41 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 27110 invoked from network); 6 Sep 2002 15:02:40 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail15.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 6 Sep 2002 15:02:40 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g86F2dBv014829; Fri, 6 Sep 2002 11:02:39 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <15735.51130.562617.753358@grasshopper.cs.duke.edu> Date: Fri, 06 Sep 2002 11:02:39 -0400 (EDT) From: John Baldwin To: Andrew Gallatin Subject: Re: alpha performance on -current Cc: freebsd-alpha@FreeBSD.ORG Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 05-Sep-2002 Andrew Gallatin wrote: > > John Baldwin writes: > > Oh, go ahead and commit, it is current after all. :) I'll cvs up and > > test it tomorrow. :) > > I'd rather not commit code if it might be broken. I'd greatly prefer > it if somebody with an MP box could try it.. > > I've left it in ~gallatin/atomic.patch on freefall, for easier > patching. Well, I'm currently doing 4.7 release test builds on the ds20, but I'll try to get to this stuff when I can. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 8: 2:48 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 001A337B401 for ; Fri, 6 Sep 2002 08:02:43 -0700 (PDT) Received: from mail.speakeasy.net (mail17.speakeasy.net [216.254.0.217]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8A95543E3B for ; Fri, 6 Sep 2002 08:02:43 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 32413 invoked from network); 6 Sep 2002 15:02:41 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail17.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 6 Sep 2002 15:02:41 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g86F2fBv014836; Fri, 6 Sep 2002 11:02:41 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <3D7806D1.C87A9A6A@mindspring.com> Date: Fri, 06 Sep 2002 11:02:41 -0400 (EDT) From: John Baldwin To: Terry Lambert Subject: Re: ithread preemption Cc: freebsd-alpha@FreeBSD.ORG, Andrew Gallatin Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 06-Sep-2002 Terry Lambert wrote: > Andrew Gallatin wrote: >> John Baldwin writes: >> > Solaris doesn't run on alpha, but it also a bit different in its approach. >> > I do wonder if there is a way we can violate an assumption in PAL due to >> > migration though. That is, a thread could return to PAL on a different >> > CPU than the one the interrupt was originally sent to. This might explain >> > why only SMP has problems. >> > >> >> Hey ... I think you have it on the nose! That makes the most sense >> I've heard yet. >> >> Do we have any way to bind a thread to a cpu? > > Alfred has some patches. That are just as tore up by KSE as mine are, probably more so. > IMO, threads should not be so bound to a CPU, as a migration event > should be an exception, rather than the rule, and always handled > by the scheduler. Yes, we all know this and plan to do this when the scheduler isn't under so much upheaval (read 6.0 when KSE finishes breaking^Wchanging things). -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 8: 2:51 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6446C37B408 for ; Fri, 6 Sep 2002 08:02:47 -0700 (PDT) Received: from mail.speakeasy.net (mail13.speakeasy.net [216.254.0.213]) by mx1.FreeBSD.org (Postfix) with ESMTP id CE84043E72 for ; Fri, 6 Sep 2002 08:02:46 -0700 (PDT) (envelope-from jhb@FreeBSD.org) Received: (qmail 7161 invoked from network); 6 Sep 2002 15:02:44 -0000 Received: from unknown (HELO server.baldwin.cx) ([216.27.160.63]) (envelope-sender ) by mail13.speakeasy.net (qmail-ldap-1.03) with DES-CBC3-SHA encrypted SMTP for ; 6 Sep 2002 15:02:45 -0000 Received: from laptop.baldwin.cx (gw1.twc.weather.com [216.133.140.1]) by server.baldwin.cx (8.12.5/8.12.5) with ESMTP id g86F2hBv014839; Fri, 6 Sep 2002 11:02:43 -0400 (EDT) (envelope-from jhb@FreeBSD.org) Message-ID: X-Mailer: XFMail 1.5.2 on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <3D78098B.CEBF13EC@mindspring.com> Date: Fri, 06 Sep 2002 11:02:43 -0400 (EDT) From: John Baldwin To: Terry Lambert Subject: Re: ithread preemption Cc: freebsd-alpha@FreeBSD.ORG, Andrew Gallatin Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On 06-Sep-2002 Terry Lambert wrote: > John Baldwin wrote: >> > Hey ... I think you have it on the nose! That makes the most sense >> > I've heard yet. >> > >> > Do we have any way to bind a thread to a cpu? >> >> I used to. Then KSE3 was committed. I suppose I could rewrite it from >> scratch again. > > The ithreads are different from threads. People don't seem to > get this idea, but it is nevertheless true. Having general support for binding threads to a CPU is something in the scheduler itself, not something buried in ithreads alone. The rest of your message is irrelevant to the actual situation at hand. -- John Baldwin <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 9:19:34 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6F85A37B400 for ; Fri, 6 Sep 2002 09:19:32 -0700 (PDT) Received: from creme-brulee.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id B4FFE43E65 for ; Fri, 6 Sep 2002 09:19:31 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from [10.2.1.0] (vpn-client-0.marcuscom.com [10.2.1.0]) by creme-brulee.marcuscom.com (8.12.5/8.12.5) with ESMTP id g86GHUix077984; Fri, 6 Sep 2002 12:17:31 -0400 (EDT) (envelope-from marcus@marcuscom.com) Subject: Re: mozilla failure on 4.6.2 From: Joe Marcus Clarke To: Andrew Gallatin Cc: alpha@FreeBSD.ORG, Jan Lentfer In-Reply-To: <15736.40749.986834.818558@grasshopper.cs.duke.edu> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 12:19:09 -0400 Message-Id: <1031329151.342.9.camel@gyros.marcuscom.com> Mime-Version: 1.0 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 2002-09-06 at 08:27, Andrew Gallatin wrote: > > Joe Marcus Clarke writes: > > On Fri, 2002-09-06 at 00:06, Kris Kennaway wrote: > > > Anyone able to take a look? > > > > > > http://beta.freebsd.org/errorlogs/4-latest/mozilla-1.0.rc3,1.log > > > > This is "fixed" now. The reason I say "fixed" is that Mozilla is > > currently broken on Alpha FreeBSD. The problem appears to be issues > > with Alpha assembly code in our version of the GCC compiler. > > Can you elaborate, please? This is the URL I found when trying to troubleshoot the problems with the seg faults: http://www.mozilla.org/unix/tru64.html It applies to Tru64, but it hints that Alpha Linux is the only platform capable of compiling Mozilla for Alpha using GCC. The Compaq C++ compiler is, unfortunately, not ported to FreeBSD. Jan Lentfer ran into another URL, that I don't have at the moment. I've copied him as perhaps he remembers it. Joe > > Thanks, > > Drew > -- PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 9:36: 6 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DAA3C37B400 for ; Fri, 6 Sep 2002 09:36:02 -0700 (PDT) Received: from srv1.cosmo-project.de (srv1.cosmo-project.de [213.83.6.106]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8AEEF43E3B for ; Fri, 6 Sep 2002 09:36:01 -0700 (PDT) (envelope-from ticso@cicely9.cicely.de) Received: from cicely5.cicely.de (cicely5.cicely.de [IPv6:3ffe:400:8d0:301:200:92ff:fe9b:20e7]) by srv1.cosmo-project.de (8.12.5/8.12.5) with ESMTP id g86GZv6K084645 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=OK); Fri, 6 Sep 2002 18:35:58 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (cicely9.cicely.de [IPv6:3ffe:400:8d0:301:210:5aff:fe30:1c1a]) by cicely5.cicely.de (8.12.6/8.12.6) with ESMTP id g86GZuNk007851 (version=TLSv1/SSLv3 cipher=EDH-RSA-DES-CBC3-SHA bits=168 verify=NO); Fri, 6 Sep 2002 18:35:56 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: from cicely9.cicely.de (localhost [127.0.0.1]) by cicely9.cicely.de (8.12.6/8.12.5) with ESMTP id g86GZTgm000946; Fri, 6 Sep 2002 18:35:29 +0200 (CEST) (envelope-from ticso@cicely9.cicely.de) Received: (from ticso@localhost) by cicely9.cicely.de (8.12.6/8.12.6/Submit) id g86GZSPg000945; Fri, 6 Sep 2002 18:35:28 +0200 (CEST) Date: Fri, 6 Sep 2002 18:35:28 +0200 From: Bernd Walter To: Joe Marcus Clarke Cc: Andrew Gallatin , alpha@FreeBSD.ORG, Jan Lentfer Subject: Re: mozilla failure on 4.6.2 Message-ID: <20020906163527.GD764@cicely9.cicely.de> Reply-To: ticso@cicely.de References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1031329151.342.9.camel@gyros.marcuscom.com> X-Operating-System: FreeBSD cicely9.cicely.de 5.0-CURRENT alpha User-Agent: Mutt/1.5.1i Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > On Fri, 2002-09-06 at 08:27, Andrew Gallatin wrote: > > > > Joe Marcus Clarke writes: > > > On Fri, 2002-09-06 at 00:06, Kris Kennaway wrote: > > > > Anyone able to take a look? > > > > > > > > http://beta.freebsd.org/errorlogs/4-latest/mozilla-1.0.rc3,1.log 1.1 is in the ports since some time now. > > > This is "fixed" now. The reason I say "fixed" is that Mozilla is > > > currently broken on Alpha FreeBSD. The problem appears to be issues > > > with Alpha assembly code in our version of the GCC compiler. > > > > Can you elaborate, please? > > This is the URL I found when trying to troubleshoot the problems with > the seg faults: > > http://www.mozilla.org/unix/tru64.html > > It applies to Tru64, but it hints that Alpha Linux is the only platform > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > compiler is, unfortunately, not ported to FreeBSD. > > Jan Lentfer ran into another URL, that I don't have > at the moment. I've copied him as perhaps he remembers it. This is much different from what I'm seeing when trying mozilla on -current for a long time. I get many missing symbols on Stub+numbers rather late in the build. If you nee to know I can start a build. I got the errors with 1.0 and I also get them with recent ports. Even if I previously remove all ports. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 10:19:19 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 5693837B400 for ; Fri, 6 Sep 2002 10:19:17 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 6BBCF43E3B for ; Fri, 6 Sep 2002 10:19:16 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id NAA12658; Fri, 6 Sep 2002 13:19:15 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86HIjR60103; Fri, 6 Sep 2002 13:18:45 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15736.58229.652949.93794@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 13:18:45 -0400 (EDT) To: Joe Marcus Clarke Cc: alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <1031329151.342.9.camel@gyros.marcuscom.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Joe Marcus Clarke writes: > http://www.mozilla.org/unix/tru64.html > > It applies to Tru64, but it hints that Alpha Linux is the only platform > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > compiler is, unfortunately, not ported to FreeBSD. And probably never will be, as it requires symbols from the linux libc, the last I remember. I assume you're talking about xptcall? From the link at the bottom, there appears to be code for linux alpha. Since we use the same g++ compiler, and the same binary format, I suspect that the linux assembly code there should work. And it appears we're using it: <1:13pm>monet/gallatin:unix>pwd /usr/ports/www/mozilla-devel/work/mozilla/xpcom/reflect/xptcall/src/md/unix <1:13pm>monet/gallatin:unix>ls *.o xptcinvoke_linux_alpha.o xptcstubs_linux_alpha.o Are you saying that there's something different about the calling conventions that would make using the linux code as it stands incorrect? Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 10:25:49 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4B4E537B400 for ; Fri, 6 Sep 2002 10:25:46 -0700 (PDT) Received: from creme-brulee.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id A975A43E65 for ; Fri, 6 Sep 2002 10:25:45 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from [10.2.1.0] (vpn-client-0.marcuscom.com [10.2.1.0]) by creme-brulee.marcuscom.com (8.12.5/8.12.5) with ESMTP id g86HNoix078416; Fri, 6 Sep 2002 13:23:50 -0400 (EDT) (envelope-from marcus@marcuscom.com) Subject: Re: mozilla failure on 4.6.2 From: Joe Marcus Clarke To: Andrew Gallatin Cc: alpha@FreeBSD.ORG In-Reply-To: <15736.58229.652949.93794@grasshopper.cs.duke.edu> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 13:25:29 -0400 Message-Id: <1031333130.342.46.camel@gyros.marcuscom.com> Mime-Version: 1.0 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 2002-09-06 at 13:18, Andrew Gallatin wrote: > > Joe Marcus Clarke writes: > > http://www.mozilla.org/unix/tru64.html > > > > It applies to Tru64, but it hints that Alpha Linux is the only platform > > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > > compiler is, unfortunately, not ported to FreeBSD. > > And probably never will be, as it requires symbols from the linux > libc, the last I remember. > > I assume you're talking about xptcall? From the link at the bottom, > there appears to be code for linux alpha. Since we use the same g++ > compiler, and the same binary format, I suspect that the linux > assembly code there should work. And it appears we're using it: > > <1:13pm>monet/gallatin:unix>pwd > /usr/ports/www/mozilla-devel/work/mozilla/xpcom/reflect/xptcall/src/md/unix > <1:13pm>monet/gallatin:unix>ls *.o > xptcinvoke_linux_alpha.o xptcstubs_linux_alpha.o > > Are you saying that there's something different about the calling > conventions that would make using the linux code as it stands > incorrect? I don't know. However, I am seeing core dumps on the port-build and when executing mozilla that point to xptcall. Admittedly, I don't have an Alpha to test with myself. Jan had been generous enough to allow me to login to his. He is running 4.6.2, so I have not been able to test -stable or -current. However, tests with gcc31 in ports, and with varying degrees of optimizations have yielded no success on Alpha. Joe > > Drew > -- PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 10:36:32 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id BD81837B400 for ; Fri, 6 Sep 2002 10:36:29 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0861843E6E for ; Fri, 6 Sep 2002 10:36:29 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id NAA13438; Fri, 6 Sep 2002 13:36:25 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86HZtX60122; Fri, 6 Sep 2002 13:35:55 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15736.59259.51892.798026@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 13:35:55 -0400 (EDT) To: Joe Marcus Clarke Cc: alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <1031333130.342.46.camel@gyros.marcuscom.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> <1031333130.342.46.camel@gyros.marcuscom.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Joe Marcus Clarke writes: > > Are you saying that there's something different about the calling > > conventions that would make using the linux code as it stands > > incorrect? > > I don't know. However, I am seeing core dumps on the port-build and > when executing mozilla that point to xptcall. Admittedly, I don't have > an Alpha to test with myself. Jan had been generous enough to allow me > to login to his. He is running 4.6.2, so I have not been able to test > -stable or -current. > > However, tests with gcc31 in ports, and with varying degrees of > optimizations have yielded no success on Alpha. > Do you know what, if any, differences there are between our g++ and the one linux uses? I'll be the first to admit that I don't know jack about g++/c++. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 11: 5:25 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B63EA37B400 for ; Fri, 6 Sep 2002 11:05:21 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 150CC43E42 for ; Fri, 6 Sep 2002 11:05:21 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id OAA14926; Fri, 6 Sep 2002 14:05:20 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86I4jc60148; Fri, 6 Sep 2002 14:04:45 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: base64 Message-ID: <15736.60989.78548.679870@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 14:04:45 -0400 (EDT) To: Joe Marcus Clarke Cc: alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <1031333130.342.46.camel@gyros.marcuscom.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> <1031333130.342.46.camel@gyros.marcuscom.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org DQpKb2UgTWFyY3VzIENsYXJrZSB3cml0ZXM6DQogPiA+IEFyZSB5b3Ugc2F5aW5nIHRoYXQg dGhlcmUncyBzb21ldGhpbmcgZGlmZmVyZW50IGFib3V0IHRoZSBjYWxsaW5nDQogPiA+IGNv bnZlbnRpb25zIHRoYXQgd291bGQgbWFrZSB1c2luZyB0aGUgbGludXggY29kZSBhcyBpdCBz dGFuZHMNCiA+ID4gaW5jb3JyZWN0Pw0KID4gDQogPiBJIGRvbid0IGtub3cuICBIb3dldmVy LCBJIGFtIHNlZWluZyBjb3JlIGR1bXBzIG9uIHRoZSBwb3J0LWJ1aWxkIGFuZA0KID4gd2hl biBleGVjdXRpbmcgbW96aWxsYSB0aGF0IHBvaW50IHRvIHhwdGNhbGwuICBBZG1pdHRlZGx5 LCBJIGRvbid0IGhhdmUNCiA+IGFuIEFscGhhIHRvIHRlc3Qgd2l0aCBteXNlbGYuICBKYW4g aGFkIGJlZW4gZ2VuZXJvdXMgZW5vdWdoIHRvIGFsbG93IG1lDQogPiB0byBsb2dpbiB0byBo aXMuICBIZSBpcyBydW5uaW5nIDQuNi4yLCBzbyBJIGhhdmUgbm90IGJlZW4gYWJsZSB0byB0 ZXN0DQogPiAtc3RhYmxlIG9yIC1jdXJyZW50Lg0KID4gDQoNCllvdSdyZSByaWdodCAtLSB4 cHRpbnZva2Ugc2VlbXMgZipja2VkOg0KDQo8MTo1NHBtPm1vbmV0L2dhbGxhdGluOnRlc3Rz PnB3ZA0KL3Vzci9wb3J0cy93d3cvbW96aWxsYS1kZXZlbC93b3JrL21vemlsbGEveHBjb20v cmVmbGVjdC94cHRjYWxsL3Rlc3RzDQo8MTo1NXBtPm1vbmV0L2dhbGxhdGluOnRlc3RzPi4v VGVzdFhQVENJbnZva2UNCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgDQpj YWxsaW5nIGRpcmVjdDoNCgkxICsgMSA9IDINCgkxTCArIDFMID0gMg0KCTIgKiAyID0gNA0K CTJMICogMkwgPSA0DQoJMSArIDIgKyAzICsgNCArIDUgKyA2ICsgNyArIDggKyA5ICsgMTAg PSA1NQ0KCTEgKyAyID0gMy4wMDAwMDBmDQoJMSArIDIgKyAzICsgNCArIDUgKyA2ICsgNyAr IDggKyA5ICsgMTAgPSA1NS4wMDAwMDANCgkxICsgMiArIDMgKyA0ICsgNSArIDYgKyA3ICsg OCArIDkgKyAxMCA9IDU1LjAwMDAwMGYNCgkxICsgMiArIDMgKyA0ICsgNSArIDYgKyA3ICsg OCArIDkgKyAxMCArIDExICsgMTIgKyAxMyArIDE0ICsxMTUgKyAxNiArIDE3ICsgMTggKyAx OSArIDIwID0gMjEwLjAwMDAwMGYNCiAgIG1vbyBjb3cgPSBtaWxrDQpjYWxsaW5nIHZpYSBp bnZva2U6DQoJMSArIDEgPSAwDQoJMUwgKyAxTCA9IDANClNlZ21lbnRhdGlvbiBmYXVsdCAo Y29yZSBkdW1wZWQpDQoNCkkgdGhpbmsgQmVybmQgaXMgcmlnaHQgdG9vLiBUaGUgYnVpbGQg c3RvcHMgYSBsaXR0bGUgZWFybGllciB0aGFuDQp0aGUgcG9zdC1idWlsZCBzZWd2IHdlJ3Zl IGJlZW4gc2VlaW5nOg0KDQpnbWFrZVs0XTogRW50ZXJpbmcgZGlyZWN0b3J5IGAvdXNyL3Bv cnRzL3d3dy9tb3ppbGxhLWRldmVsL3dvcmsvbW96aWxsYS9hY2Nlc3NpYmxlL3NyYy9iYXNl Jw0KbnNBY2Nlc3NpYmxlLmNwcA0KYysrIC1vIG5zQWNjZXNzaWJsZS5vIC1jIC1ET1NUWVBF PVwiRnJlZUJTRDRcIiAtRE9TQVJDSD1cIkZyZWVCU0RcIiAtRE9KSSAtSS4gLUkuLy4uL2h0 bWwgLUkuLy4uL3h1bCAtSS4vLi4vLi4vLi4vbGF5b3V0L2h0bWwvYmFzZS9zcmMgIC1JLi4v Li4vLi4vZGlzdC9pbmNsdWRlL2NvbnRlbnQgLUkuLi8uLi8uLi9kaXN0L2luY2x1ZGUvZG9j c2hlbGwgLUkuLi8uLi8uLi9kaXN0L2luY2x1ZGUvZG9tIC1JLi4vLi4vLi4vZGlzdC9pbmNs dWRlL2dmeCAtSS4uLy4uLy4uL2Rpc3QvaW5jbHVkZS9odG1scGFyc2VyIC1JLi4vLi4vLi4v ZGlzdC9pbmNsdWRlL2ludGwgLUkuLi8uLi8uLi9kaXN0L2luY2x1ZGUvbGF5b3V0IC1JLi4v Li4vLi4vZGlzdC9pbmNsdWRlL2xvY2FsZSAtSS4uLy4uLy4uL2Rpc3QvaW5jbHVkZS9uZWNr byAtSS4uLy4uLy4uL2Rpc3QvaW5jbHVkZS9wbHVnaW4gLUkuLi8uLi8uLi9kaXN0L2luY2x1 ZGUvcHJlZiAtSS4uLy4uLy4uL2Rpc3QvaW5jbHVkZS9zdHJpbmcgLUkuLi8uLi8uLi9kaXN0 L2luY2x1ZGUvdXJpbG9hZGVyIC1JLi4vLi4vLi4vZGlzdC9pbmNsdWRlL3ZpZXcgLUkuLi8u Li8uLi9kaXN0L2luY2x1ZGUvd2Vic2hlbGwgLUkuLi8uLi8uLi9kaXN0L2luY2x1ZGUvd2lk Z2V0IC1JLi4vLi4vLi4vZGlzdC9pbmNsdWRlL3hwY29tIC1JLi4vLi4vLi4vZGlzdC9pbmNs dWRlL3h1bGRvYyAtSS4uLy4uLy4uL2Rpc3QvaW5jbHVkZS9hY2Nlc3NpYmlsaXR5IC1JLi4v Li4vLi4vZGlzdC9pbmNsdWRlIC1JL3Vzci9wb3J0cy93d3cvbW96aWxsYS1kZXZlbC93b3Jr L21vemlsbGEvZGlzdC9pbmNsdWRlL25zcHIgIC1JL3Vzci9sb2NhbC9pbmNsdWRlIC1JL3Vz ci9sb2NhbC9pbmNsdWRlICAgLUkvdXNyL1gxMVI2L2luY2x1ZGUgICAtZlBJQyAgLUkvdXNy L1gxMVI2L2luY2x1ZGUgIC1JL3Vzci9YMTFSNi9pbmNsdWRlIC1mbm8tcnR0aSAtZm5vLWV4 Y2VwdGlvbnMgLVdhbGwgLVdjb252ZXJzaW9uIC1XcG9pbnRlci1hcml0aCAtV2JhZC1mdW5j dGlvbi1jYXN0IC1XY2FzdC1hbGlnbiAtV292ZXJsb2FkZWQtdmlydHVhbCAtV3N5bnRoIC1X bm8tY3Rvci1kdG9yLXByaXZhY3kgLVduby1sb25nLWxvbmcgLU8gLXBpcGUgLW1jcHU9ZXY1 NiAtcGlwZSAgLUROREVCVUcgLURUUklNTUVEICAtSS91c3IvWDExUjYvaW5jbHVkZSAgLUkv dXNyL1gxMVI2L2luY2x1ZGUgLURNT1pJTExBX0NMSUVOVCAtaW5jbHVkZSAuLi8uLi8uLi9j b25maWctZGVmcy5oIG5zQWNjZXNzaWJsZS5jcHANCkluIGZpbGUgaW5jbHVkZWQgZnJvbSBu c0FjY2Vzc2libGUuY3BwOjQxOg0KbnNBY2Nlc3NpYmxlLmg6MTc2OiB3YXJuaW5nOiBgY2xh c3MgbnNBY2Nlc3NpYmxlVHJlZVdhbGtlcicgaGFzIHZpcnR1YWwgZnVuY3Rpb25zIGJ1dCBu b24tdmlydHVhbCBkZXN0cnVjdG9yDQpuc0FjY2Vzc2libGUuY3BwOiBJbiBtZXRob2QgYG5z cmVzdWx0IG5zQWNjZXNzaWJsZTo6R2V0QWNjSWQoUFJJbnQzMiAqKSc6DQpuc0FjY2Vzc2li bGUuY3BwOjUzMDogcmVpbnRlcnByZXRfY2FzdCBmcm9tIGBuc0Rlcml2ZWRTYWZlPG5zSURP TU5vZGU+IConIHRvIGBQUkludDMyJyBsb3NlcyBwcmVjaXNpb24NCm5zQWNjZXNzaWJsZS5j cHA6IEluIG1ldGhvZCBgbnNyZXN1bHQgbnNBY2Nlc3NpYmxlOjpBcHBlbmRGbGF0U3RyaW5n RnJvbUNvbnRlbnROb2RlKG5zSUNvbnRlbnQgKiwgbnNBU3RyaW5nICopJzoNCm5zQWNjZXNz aWJsZS5jcHA6MTI3NDogd2FybmluZzogY2FzdCBpbmNyZWFzZXMgcmVxdWlyZWQgYWxpZ25t ZW50IG9mIHRhcmdldCB0eXBlDQpnbWFrZVs0XTogKioqIFtuc0FjY2Vzc2libGUub10gRXJy b3IgMQ0KZ21ha2VbNF06IExlYXZpbmcgZGlyZWN0b3J5IGAvdXNyL3BvcnRzL3d3dy9tb3pp bGxhLWRldmVsL3dvcmsvbW96aWxsYS9hY2Nlc3NpYmxlL3NyYy9iYXNlJw0KZ21ha2VbM106 ICoqKiBbbGlic10gRXJyb3IgMg0KZ21ha2VbM106IExlYXZpbmcgZGlyZWN0b3J5IGAvdXNy L3BvcnRzL3d3dy9tb3ppbGxhLWRldmVsL3dvcmsvbW96aWxsYS9hY2Nlc3NpYmxlL3NyYycN CmdtYWtlWzJdOiAqKiogW2xpYnNdIEVycm9yIDINCmdtYWtlWzJdOiBMZWF2aW5nIGRpcmVj dG9yeSBgL3Vzci9wb3J0cy93d3cvbW96aWxsYS1kZXZlbC93b3JrL21vemlsbGEvYWNjZXNz aWJsZScNCmdtYWtlWzFdOiAqKiogW3RpZXJfOV0gRXJyb3IgMg0KZ21ha2VbMV06IExlYXZp bmcgZGlyZWN0b3J5IGAvdXNyL3BvcnRzL3d3dy9tb3ppbGxhLWRldmVsL3dvcmsvbW96aWxs YScNCmdtYWtlOiAqKiogW2RlZmF1bHRdIEVycm9yIDINCioqKiBFcnJvciBjb2RlIDINCg0K RHJldw== To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 11:33:42 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1874737B400 for ; Fri, 6 Sep 2002 11:33:40 -0700 (PDT) Received: from creme-brulee.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 731B443E4A for ; Fri, 6 Sep 2002 11:33:39 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from [10.2.1.0] (vpn-client-0.marcuscom.com [10.2.1.0]) by creme-brulee.marcuscom.com (8.12.5/8.12.5) with ESMTP id g86IViix078818; Fri, 6 Sep 2002 14:31:44 -0400 (EDT) (envelope-from marcus@marcuscom.com) Subject: Re: mozilla failure on 4.6.2 From: Joe Marcus Clarke To: Andrew Gallatin Cc: alpha@FreeBSD.ORG In-Reply-To: <15736.59259.51892.798026@grasshopper.cs.duke.edu> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> <1031333130.342.46.camel@gyros.marcuscom.com> <15736.59259.51892.798026@grasshopper.cs.duke.edu> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 14:33:23 -0400 Message-Id: <1031337204.342.49.camel@gyros.marcuscom.com> Mime-Version: 1.0 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 2002-09-06 at 13:35, Andrew Gallatin wrote: > > Joe Marcus Clarke writes: > > > Are you saying that there's something different about the calling > > > conventions that would make using the linux code as it stands > > > incorrect? > > > > I don't know. However, I am seeing core dumps on the port-build and > > when executing mozilla that point to xptcall. Admittedly, I don't have > > an Alpha to test with myself. Jan had been generous enough to allow me > > to login to his. He is running 4.6.2, so I have not been able to test > > -stable or -current. > > > > However, tests with gcc31 in ports, and with varying degrees of > > optimizations have yielded no success on Alpha. > > > > Do you know what, if any, differences there are between our g++ and > the one linux uses? I'll be the first to admit that I don't know jack > about g++/c++. Sorry, neither do I. However, the builds of Mozilla I've seen for Linux Alpha use GCC 2.96 where as I've been testing with 2.95.x. Joe > > Drew > -- PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 12: 2: 2 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7AAD937B400 for ; Fri, 6 Sep 2002 12:02:00 -0700 (PDT) Received: from smtp.web.de (smtp03.web.de [217.72.192.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 7FEC743E3B for ; Fri, 6 Sep 2002 12:01:59 -0700 (PDT) (envelope-from Jan.Lentfer@web.de) Received: from [217.82.32.240] (helo=floundjan.homeip.net) by smtp.web.de with esmtp (WEB.DE(Exim) 4.75 #2) id 17nOMB-0002O9-00; Fri, 06 Sep 2002 21:01:27 +0200 Received: from localhost (localhost.lan [127.0.0.1]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id 80D3D29D; Fri, 6 Sep 2002 21:01:24 +0200 (CEST) Received: from jan-linux.lan (jan-linux.lan [192.168.0.20]) by floundjan.homeip.net (Postfix on FreeBSD 4.5) with ESMTP id E355F28F; Fri, 6 Sep 2002 21:01:19 +0200 (CEST) Subject: Re: mozilla failure on 4.6.2 From: Jan Lentfer To: Joe Marcus Clarke Cc: Andrew Gallatin , alpha@FreeBSD.ORG In-Reply-To: <1031337204.342.49.camel@gyros.marcuscom.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> <1031333130.342.46.camel@gyros.marcuscom.com> <15736.59259.51892.798026@grasshopper.cs.duke.edu> <1031337204.342.49.camel@gyros.marcuscom.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.7-1mdk Date: 06 Sep 2002 21:01:17 +0200 Message-Id: <1031338877.5209.2.camel@jan-linux.lan> Mime-Version: 1.0 X-Virus-Scanned: by AMaViS perl-10 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Am Fre, 2002-09-06 um 20.33 schrieb Joe Marcus Clarke: > On Fri, 2002-09-06 at 13:35, Andrew Gallatin wrote: > > Do you know what, if any, differences there are between our g++ and > > the one linux uses? I'll be the first to admit that I don't know jack > > about g++/c++. > > Sorry, neither do I. However, the builds of Mozilla I've seen for Linux > Alpha use GCC 2.96 where as I've been testing with 2.95.x. Joe, I think wie also tried with gcc31 and both with different optimization levels (-O & -O0), didn't we? All failed at least :-( Jan -- Jan Lentfer System Administrator Molecular Cell Biology / AG Holstein, Darmstadt University of Technology, Schnittspahnstr. 10, 64287 Darmstadt, Germany Tel: +49 6151 16 5563 / Tel private: +49 6151 788415 / mobile: +49 163 4712037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 12:15:36 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 5124037B401 for ; Fri, 6 Sep 2002 12:15:32 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 8810C43E6A for ; Fri, 6 Sep 2002 12:15:31 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id PAA18464; Fri, 6 Sep 2002 15:15:30 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86JF0F60208; Fri, 6 Sep 2002 15:15:00 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15736.65204.564404.552876@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 15:15:00 -0400 (EDT) To: Jan Lentfer Cc: Joe Marcus Clarke , alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <1031338877.5209.2.camel@jan-linux.lan> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <15736.58229.652949.93794@grasshopper.cs.duke.edu> <1031333130.342.46.camel@gyros.marcuscom.com> <15736.59259.51892.798026@grasshopper.cs.duke.edu> <1031337204.342.49.camel@gyros.marcuscom.com> <1031338877.5209.2.camel@jan-linux.lan> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Jan Lentfer writes: > > > Do you know what, if any, differences there are between our g++ and > > > the one linux uses? I'll be the first to admit that I don't know jack > > > about g++/c++. > > > > Sorry, neither do I. However, the builds of Mozilla I've seen for Linux > > Alpha use GCC 2.96 where as I've been testing with 2.95.x. > > Joe, > > I think wie also tried with gcc31 and both with different optimization > levels (-O & -O0), didn't we? All failed at least :-( None of that matters if the calling conventions are different. Some quick debugging reveals we're jumping off into never-never land, rather than really calling the function which is supposed to be invoked. Eg, we end up in: InvokeTestTarget::InvokeTestTarget() (which contains no code, afaik), rather than InvokeTestTarget::AddTwoInts It might be as simple as adding an offset to something, but I'm about out of time for today. My lunch hour only lasts but so long .. ;) Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 12:47:45 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EB11B37B407 for ; Fri, 6 Sep 2002 12:47:37 -0700 (PDT) Received: from dragon.nuxi.com (trang.nuxi.com [66.92.13.169]) by mx1.FreeBSD.org (Postfix) with ESMTP id 29E4F43E42 for ; Fri, 6 Sep 2002 12:47:37 -0700 (PDT) (envelope-from obrien@NUXI.com) Received: from dragon.nuxi.com (obrien@localhost [127.0.0.1]) by dragon.nuxi.com (8.12.6/8.12.2) with ESMTP id g86JlXfO002340; Fri, 6 Sep 2002 12:47:33 -0700 (PDT) (envelope-from obrien@dragon.nuxi.com) Received: (from obrien@localhost) by dragon.nuxi.com (8.12.6/8.12.5/Submit) id g86JlWTo002339; Fri, 6 Sep 2002 12:47:32 -0700 (PDT) Date: Fri, 6 Sep 2002 12:47:32 -0700 From: "David O'Brien" To: Joe Marcus Clarke Cc: Andrew Gallatin , alpha@FreeBSD.ORG, Jan Lentfer Subject: Re: mozilla failure on 4.6.2 Message-ID: <20020906194732.GA2244@dragon.nuxi.com> Reply-To: obrien@FreeBSD.ORG Mail-Followup-To: David O'Brien , Joe Marcus Clarke , Andrew Gallatin , alpha@FreeBSD.ORG, Jan Lentfer References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1031329151.342.9.camel@gyros.marcuscom.com> User-Agent: Mutt/1.4i X-Operating-System: FreeBSD 5.0-CURRENT Organization: The NUXI BSD Group X-Pgp-Rsa-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Rsa-Keyid: 1024/34F9F9D5 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > http://www.mozilla.org/unix/tru64.html > > It applies to Tru64, but it hints that Alpha Linux is the only platform > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > compiler is, unfortunately, not ported to FreeBSD. 2.There is xptcall code for g++ running under Linux/alpha. This code would be a good place to start writing a Tru64 implementation. The vtable format is different between g++ running under Linux and Tru64, This should not be the case for FreeBSD -- meaning we are using the same vtable format as AlphaLinux now. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 12:50:50 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 1CB2D37B400; Fri, 6 Sep 2002 12:50:48 -0700 (PDT) Received: from creme-brulee.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id 62F9843E65; Fri, 6 Sep 2002 12:50:47 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from [10.2.1.0] (vpn-client-0.marcuscom.com [10.2.1.0]) by creme-brulee.marcuscom.com (8.12.5/8.12.5) with ESMTP id g86Jmlix079267; Fri, 6 Sep 2002 15:48:47 -0400 (EDT) (envelope-from marcus@marcuscom.com) Subject: Re: mozilla failure on 4.6.2 From: Joe Marcus Clarke To: obrien@FreeBSD.ORG Cc: Andrew Gallatin , alpha@FreeBSD.ORG, Jan Lentfer In-Reply-To: <20020906194732.GA2244@dragon.nuxi.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <20020906194732.GA2244@dragon.nuxi.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 06 Sep 2002 15:50:26 -0400 Message-Id: <1031341827.342.64.camel@gyros.marcuscom.com> Mime-Version: 1.0 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 2002-09-06 at 15:47, David O'Brien wrote: > On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > > http://www.mozilla.org/unix/tru64.html > > > > It applies to Tru64, but it hints that Alpha Linux is the only platform > > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > > compiler is, unfortunately, not ported to FreeBSD. > > 2.There is xptcall code for g++ running under Linux/alpha. This code > would be a good place to start writing a Tru64 implementation. The > vtable format is different between g++ running under Linux and Tru64, > > This should not be the case for FreeBSD -- meaning we are using the same > vtable format as AlphaLinux now. As of when? Or has this always been the case? Joe > -- PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 13: 3:43 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 53A3D37B447; Fri, 6 Sep 2002 13:03:34 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 57B6843E4A; Fri, 6 Sep 2002 13:03:33 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id QAA20629; Fri, 6 Sep 2002 16:03:32 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g86K32C60276; Fri, 6 Sep 2002 16:03:02 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15737.2550.488154.4401@grasshopper.cs.duke.edu> Date: Fri, 6 Sep 2002 16:03:02 -0400 (EDT) To: obrien@FreeBSD.ORG Cc: Joe Marcus Clarke , alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <20020906194732.GA2244@dragon.nuxi.com> References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <20020906194732.GA2244@dragon.nuxi.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org David O'Brien writes: > On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > > http://www.mozilla.org/unix/tru64.html > > > > It applies to Tru64, but it hints that Alpha Linux is the only platform > > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > > compiler is, unfortunately, not ported to FreeBSD. > > 2.There is xptcall code for g++ running under Linux/alpha. This code > would be a good place to start writing a Tru64 implementation. The > vtable format is different between g++ running under Linux and Tru64, > > This should not be the case for FreeBSD -- meaning we are using the same > vtable format as AlphaLinux now. You're probably still catching up.. but my tests using TestXPTCInvoke.cpp from /usr/ports/www/mozilla-devel/work/mozilla/xpcom/reflect/xptcall/tests indicates that the linux/alpha code does not work for us. This was tested on -stable, built as of last night. It seems to be close -- the PC ends up *close* to where it should, but it does not end up at the correct address, and all hell breaks loose. Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 17: 9:58 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2730937B400 for ; Fri, 6 Sep 2002 17:09:56 -0700 (PDT) Received: from dragon.nuxi.com (trang.nuxi.com [66.92.13.169]) by mx1.FreeBSD.org (Postfix) with ESMTP id 66F2343EB1 for ; Fri, 6 Sep 2002 17:09:33 -0700 (PDT) (envelope-from obrien@NUXI.com) Received: from dragon.nuxi.com (obrien@localhost [127.0.0.1]) by dragon.nuxi.com (8.12.6/8.12.2) with ESMTP id g8708PfO004148; Fri, 6 Sep 2002 17:08:25 -0700 (PDT) (envelope-from obrien@dragon.nuxi.com) Received: (from obrien@localhost) by dragon.nuxi.com (8.12.6/8.12.5/Submit) id g8708PRs004147; Fri, 6 Sep 2002 17:08:25 -0700 (PDT) Date: Fri, 6 Sep 2002 17:08:25 -0700 From: "David O'Brien" To: Andrew Gallatin Cc: Joe Marcus Clarke , alpha@FreeBSD.ORG Subject: Re: mozilla failure on 4.6.2 Message-ID: <20020907000825.GA4091@dragon.nuxi.com> Reply-To: obrien@FreeBSD.ORG Mail-Followup-To: David O'Brien , Andrew Gallatin , Joe Marcus Clarke , alpha@FreeBSD.ORG References: <20020906040641.GA61111@xor.obsecurity.org> <1031287598.361.2.camel@gyros.marcuscom.com> <15736.40749.986834.818558@grasshopper.cs.duke.edu> <1031329151.342.9.camel@gyros.marcuscom.com> <20020906194732.GA2244@dragon.nuxi.com> <15737.2550.488154.4401@grasshopper.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <15737.2550.488154.4401@grasshopper.cs.duke.edu> User-Agent: Mutt/1.4i X-Operating-System: FreeBSD 5.0-CURRENT Organization: The NUXI BSD Group X-Pgp-Rsa-Fingerprint: B7 4D 3E E9 11 39 5F A3 90 76 5D 69 58 D9 98 7A X-Pgp-Rsa-Keyid: 1024/34F9F9D5 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, Sep 06, 2002 at 04:03:02PM -0400, Andrew Gallatin wrote: > > On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > > > http://www.mozilla.org/unix/tru64.html > > > > > > It applies to Tru64, but it hints that Alpha Linux is the only platform > > > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > > > compiler is, unfortunately, not ported to FreeBSD. > > > > 2.There is xptcall code for g++ running under Linux/alpha. This code > > would be a good place to start writing a Tru64 implementation. The > > vtable format is different between g++ running under Linux and Tru64, > > > > This should not be the case for FreeBSD -- meaning we are using the same > > vtable format as AlphaLinux now. > > You're probably still catching up.. but my tests using TestXPTCInvoke.cpp > from /usr/ports/www/mozilla-devel/work/mozilla/xpcom/reflect/xptcall/tests > indicates that the linux/alpha code does not work for us. > > This was tested on -stable, built as of last night. The vtable format in stable is not the same as AlphaLinux, it is the same in -CURRENT. Do we know if the results you posted also happen on -CURRENT? To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Fri Sep 6 22:56:10 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8B2FA37B400; Fri, 6 Sep 2002 22:56:08 -0700 (PDT) Received: from shumai.marcuscom.com (rdu57-17-158.nc.rr.com [66.57.17.158]) by mx1.FreeBSD.org (Postfix) with ESMTP id EC94B43E65; Fri, 6 Sep 2002 22:56:07 -0700 (PDT) (envelope-from marcus@marcuscom.com) Received: from shumai.marcuscom.com (localhost.marcuscom.com [127.0.0.1]) by shumai.marcuscom.com (8.12.5/8.12.5) with ESMTP id g875uRVD045981; Sat, 7 Sep 2002 01:56:27 -0400 (EDT) (envelope-from marcus@marcuscom.com) Received: from localhost (marcus@localhost) by shumai.marcuscom.com (8.12.5/8.12.5/Submit) with ESMTP id g875uRgj045978; Sat, 7 Sep 2002 01:56:27 -0400 (EDT) X-Authentication-Warning: shumai.marcuscom.com: marcus owned process doing -bs Date: Sat, 7 Sep 2002 01:56:27 -0400 (EDT) From: Joe Marcus Clarke To: "David O'Brien" Cc: Andrew Gallatin , Subject: Re: mozilla failure on 4.6.2 In-Reply-To: <20020907000825.GA4091@dragon.nuxi.com> Message-ID: <20020907015556.L45915-100000@shumai.marcuscom.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Fri, 6 Sep 2002, David O'Brien wrote: > On Fri, Sep 06, 2002 at 04:03:02PM -0400, Andrew Gallatin wrote: > > > On Fri, Sep 06, 2002 at 12:19:09PM -0400, Joe Marcus Clarke wrote: > > > > http://www.mozilla.org/unix/tru64.html > > > > > > > > It applies to Tru64, but it hints that Alpha Linux is the only platform > > > > capable of compiling Mozilla for Alpha using GCC. The Compaq C++ > > > > compiler is, unfortunately, not ported to FreeBSD. > > > > > > 2.There is xptcall code for g++ running under Linux/alpha. This code > > > would be a good place to start writing a Tru64 implementation. The > > > vtable format is different between g++ running under Linux and Tru64, > > > > > > This should not be the case for FreeBSD -- meaning we are using the same > > > vtable format as AlphaLinux now. > > > > You're probably still catching up.. but my tests using TestXPTCInvoke.cpp > > from /usr/ports/www/mozilla-devel/work/mozilla/xpcom/reflect/xptcall/tests > > indicates that the linux/alpha code does not work for us. > > > > This was tested on -stable, built as of last night. > > The vtable format in stable is not the same as AlphaLinux, it is the same > in -CURRENT. Do we know if the results you posted also happen on > -CURRENT? Apparently, -CURRENT has some other issues. Due to my lack of Alpha hardware, I have only done limited testing on 4.6.2-RELEASE. Joe > > PGP Key : http://www.marcuscom.com/pgp.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sat Sep 7 17:40: 1 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8A94E37B400 for ; Sat, 7 Sep 2002 17:39:55 -0700 (PDT) Received: from mail.redlinenetworks.com (mail.redlinenetworks.com [216.136.145.172]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2BE9E43E42 for ; Sat, 7 Sep 2002 17:39:55 -0700 (PDT) (envelope-from sewall@ix.netcom.com) Received: from ix.netcom.com (melkor.redlinenetworks.com [192.168.0.8]) by mail.redlinenetworks.com (8.11.6/8.11.1) with ESMTP id g880dqK43212; Sat, 7 Sep 2002 17:39:54 -0700 (PDT) (envelope-from sewall@ix.netcom.com) Message-ID: <3D7A9E7C.5020604@ix.netcom.com> Date: Sat, 07 Sep 2002 17:49:00 -0700 From: Scott Sewall User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4.1) Gecko/20020508 Netscape6/6.2.3 X-Accept-Language: en-us MIME-Version: 1.0 To: freebsd-alpha@FreeBSD.ORG Cc: sewall@ix.netcom.com Subject: Need help installing Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org I have a API CS20D, and I'm having trouble installing FreeBSD 4.6. Any help would be greatly appreciated. I'm attempting to install from a CD, made from the ISO I downloaded from freebsd.org. The install hangs after jumping to the bootstrap code: P00>>>boot dqb0 (boot dqb0.0.1.16.0 -flags 0) block 0 of dqb0.0.1.16.0 is a valid boot block reading 375 blocks from dqb0.0.1.16.0 bootstrap code read in base = 200000, image_start = 0, image_bytes = 2ee00 initializing HWRPB at 2000 initializing page table at 3ff4c000 initializing machine state setting affinity to the primary CPU jumping to bootstrap code << at this point the light on the CD blinks for several seconds >> Some other information that might be usefull: P00>>>show dev dka0.0.0.3.0 DKA0 IBM DDYS-T36950M S96H dqb0.0.1.16.0 DQB0 SAMSUNG CD-ROM SN-124 q008 eia0.0.0.3.1 EIA0 00-02-56-00-0A-36 eib0.0.0.4.0 EIB0 00-02-56-00-0A-35 pka0.7.0.3.0 PKA0 SCSI Bus ID 7 P00>>>show config API CS20D 833 MHz SRM Console: A5.8-81 PALcode: OpenVMS PALcode V1.84-66, Tru64 UNIX PALcode V1.79-64 Processors CPU 0 Alpha 21264B-4 833 MHz SROM Revision: V2.5.2.2.32 Bcache size: 4 MB CPU 1 Alpha 21264B-4 833 MHz SROM Revision: V2.5.2.2.32 Bcache size: 4 MB Core Logic Cchip DECchip 21272-CA Rev 2.1 Dchip DECchip 21272-DA Rev 2.0 Pchip 0 DECchip 21272-EA Rev 2.2 Pchip 1 DECchip 21272-EA Rev 2.2 TIG Rev 4.15 Arbiter Rev 2.8 (0x1) CS20 System Motherboard Rev 0 Power EPLD Rev 5 Reset EPLD Rev 0 MEMORY Array # Size Base Addr ------- ---------- --------- 0 1024 MB 000000000 1 1024 MB 040000000 Total Bad Pages = 0 Total Good Memory = 2048 MBytes PCI Hose 00 Bus 00 Slot 03: SYM 53C1000 pka0.7.0.3.0 SCSI Bus ID 7 dka0.0.0.3.0 IBM DDYS-T36950M Bus 00 Slot 04: Intel 8255x Ethernet eib0.0.0.4.0 00-02-56-00-0A-35 Bus 00 Slot 07: Acer Labs M1543C Bridge to Bus 1, ISA Bus 00 Slot 16: Acer Labs M1543C IDE dqa.0.0.16.0 dqb0.0.1.16.0 SAMSUNG CD-ROM SN-12 Bus 00 Slot 17: Acer Labs M1543C PMU PCI Hose 01 Bus 00 Slot 03: Intel 8255x Ethernet eia0.0.0.3.1 00-02-56-00-0A-36 ISA Slot Device Name Type Enabled BaseAddr IRQ DMA 0 0 MOUSE Embedded Yes 60 12 1 KBD Embedded Yes 60 1 2 COM1 Embedded Yes 3f8 4 3 COM2 Embedded Yes 2f8 3 4 LPT1 Embedded Yes 3bc 7 5 FLOPPY Embedded Yes 3f0 6 2 6 EIDE Embedded Yes 1f0 14 3f6 15 170 376 7 PWR_MANAGEMENT Embedded Yes 8 USB Embedded No To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message From owner-freebsd-alpha Sat Sep 7 18: 4:31 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 329B437B400 for ; Sat, 7 Sep 2002 18:04:26 -0700 (PDT) Received: from mail.redlinenetworks.com (mail.redlinenetworks.com [216.136.145.172]) by mx1.FreeBSD.org (Postfix) with ESMTP id D376643E75 for ; Sat, 7 Sep 2002 18:04:25 -0700 (PDT) (envelope-from sewall@ix.netcom.com) Received: from ix.netcom.com (melkor.redlinenetworks.com [192.168.0.8]) by mail.redlinenetworks.com (8.11.6/8.11.1) with ESMTP id g8814PK43976; Sat, 7 Sep 2002 18:04:25 -0700 (PDT) (envelope-from sewall@ix.netcom.com) Message-ID: <3D7AA43D.9010309@ix.netcom.com> Date: Sat, 07 Sep 2002 18:13:33 -0700 From: Scott Sewall User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.4.1) Gecko/20020508 Netscape6/6.2.3 X-Accept-Language: en-us MIME-Version: 1.0 To: freebsd-alpha@FreeBSD.ORG Cc: sewall@ix.netcom.com Subject: Need help installing Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org My apologies for the formatting of my last post . The embedded tabs made it very difficult to read. I have a API CS20D, and I'm having trouble installing FreeBSD 4.6. Any help would be greatly appreciated. I'm attempting to install from a CD, made from the ISO I downloaded from freebsd.org. The install hangs after jumping to the bootstrap code: P00>>>boot dqb0 (boot dqb0.0.1.16.0 -flags 0) block 0 of dqb0.0.1.16.0 is a valid boot block reading 375 blocks from dqb0.0.1.16.0 bootstrap code read in base = 200000, image_start = 0, image_bytes = 2ee00 initializing HWRPB at 2000 initializing page table at 3ff4c000 initializing machine state setting affinity to the primary CPU jumping to bootstrap code << at this point the light on the CD blinks for several seconds >> Some other information that might be usefull: P00>>>show dev dka0.0.0.3.0 DKA0 IBM DDYS-T36950M S96H dqb0.0.1.16.0 DQB0 SAMSUNG CD-ROM SN-124 q008 eia0.0.0.3.1 EIA0 00-02-56-00-0A-36 eib0.0.0.4.0 EIB0 00-02-56-00-0A-35 pka0.7.0.3.0 PKA0 SCSI Bus ID 7 P00>>>show config API CS20D 833 MHz SRM Console: A5.8-81 PALcode: OpenVMS PALcode V1.84-66, Tru64 UNIX PALcode V1.79-64 Processors CPU 0 Alpha 21264B-4 833 MHz SROM Revision: V2.5.2.2.32 Bcache size: 4 MB CPU 1 Alpha 21264B-4 833 MHz SROM Revision: V2.5.2.2.32 Bcache size: 4 MB Core Logic Cchip DECchip 21272-CA Rev 2.1 Dchip DECchip 21272-DA Rev 2.0 Pchip 0 DECchip 21272-EA Rev 2.2 Pchip 1 DECchip 21272-EA Rev 2.2 TIG Rev 4.15 Arbiter Rev 2.8 (0x1) CS20 System Motherboard Rev 0 Power EPLD Rev 5 Reset EPLD Rev 0 MEMORY Array # Size Base Addr ------- ---------- --------- 0 1024 MB 000000000 1 1024 MB 040000000 Total Bad Pages = 0 Total Good Memory = 2048 MBytes PCI Hose 00 Bus 00 Slot 03: SYM 53C1000 pka0.7.0.3.0 SCSI Bus ID 7 dka0.0.0.3.0 IBM DDYS-T36950M Bus 00 Slot 04: Intel 8255x Ethernet eib0.0.0.4.0 00-02-56-00-0A-35 Bus 00 Slot 07: Acer Labs M1543C Bridge to Bus 1, ISA Bus 00 Slot 16: Acer Labs M1543C IDE dqa.0.0.16.0 dqb0.0.1.16.0 SAMSUNG CD-ROM SN-12 Bus 00 Slot 17: Acer Labs M1543C PMU PCI Hose 01 Bus 00 Slot 03: Intel 8255x Ethernet eia0.0.0.3.1 00-02-56-00-0A-36 ISA Slot Device Name Type Enabled BaseAddr IRQ DMA 0 0 MOUSE Embedded Yes 60 12 1 KBD Embedded Yes 60 1 2 COM1 Embedded Yes 3f8 4 3 COM2 Embedded Yes 2f8 3 4 LPT1 Embedded Yes 3bc 7 5 FLOPPY Embedded Yes 3f0 6 2 6 EIDE Embedded Yes 1f0 14 3f6 15 170 376 7 PWR_MANAGEMENT Embedded Yes 8 USB Embedded No To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message