From owner-freebsd-smp Sun Sep 7 13:39:37 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA10000 for smp-outgoing; Sun, 7 Sep 1997 13:39:37 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id NAA09907; Sun, 7 Sep 1997 13:37:42 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id OAA27121; Sun, 7 Sep 1997 14:37:40 -0600 (MDT) Message-Id: <199709072037.OAA27121@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: smp@freebsd.org cc: cvs-committers@freebsd.org, cvs-all@freebsd.org, cvs-CVSROOT@freebsd.org Subject: machine/up.h Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 07 Sep 1997 14:37:40 -0600 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I propose that we create an alternate file to smp.h named up.h. code could then look like: #ifdef SMP #include #else #include #endif I could then move stuff like: #ifndef SMP #define THIS_LOCK /* nop this out for UP */ #define THAT_FEATURE /* nop this out for UP */ #endif from individual files and gather them all up into up.h This should cause less disturbance to the source tree than the current methods. Any objections? -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sun Sep 7 15:38:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA15129 for smp-outgoing; Sun, 7 Sep 1997 15:38:14 -0700 (PDT) Received: from usr04.primenet.com (tlambert@usr04.primenet.com [206.165.6.204]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id PAA14992; Sun, 7 Sep 1997 15:36:09 -0700 (PDT) Received: (from tlambert@localhost) by usr04.primenet.com (8.8.5/8.8.5) id PAA22637; Sun, 7 Sep 1997 15:36:08 -0700 (MST) From: Terry Lambert Message-Id: <199709072236.PAA22637@usr04.primenet.com> Subject: Re: machine/up.h To: smp@csn.net (Steve Passe) Date: Sun, 7 Sep 1997 22:36:07 +0000 (GMT) Cc: smp@freebsd.org, cvs-committers@freebsd.org, cvs-all@freebsd.org, cvs-CVSROOT@freebsd.org In-Reply-To: <199709072037.OAA27121@Ilsa.StevesCafe.com> from "Steve Passe" at Sep 7, 97 02:37:40 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > I propose that we create an alternate file to smp.h named up.h. > > code could then look like: > > #ifdef SMP > #include > #else > #include > #endif > > I could then move stuff like: > > #ifndef SMP > #define THIS_LOCK /* nop this out for UP */ > #define THAT_FEATURE /* nop this out for UP */ > #endif > > from individual files and gather them all up into up.h This should cause less > disturbance to the source tree than the current methods. > > Any objections? Or you could include smp.h always, and do this in smp.h: #ifdef SMP #define THIS_LOCK smp_thislock /* or whatever*/ ... #else /* !SMP*/ #define THIS_LOCK /* NOP this out for non-SMP* #endif And have even less impact. Or you could make the standard headers do the work for you, and not include smp.h at all. 8-). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-smp Sun Sep 7 16:28:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA17352 for smp-outgoing; Sun, 7 Sep 1997 16:28:03 -0700 (PDT) Received: (from jmb@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA17328; Sun, 7 Sep 1997 16:27:59 -0700 (PDT) From: "Jonathan M. Bresler" Message-Id: <199709072327.QAA17328@hub.freebsd.org> Subject: Re: machine/up.h To: smp@csn.net (Steve Passe) Date: Sun, 7 Sep 1997 16:27:58 -0700 (PDT) Cc: smp@freebsd.org, cvs-committers@freebsd.org, cvs-all@freebsd.org, cvs-CVSROOT@freebsd.org In-Reply-To: <199709072037.OAA27121@Ilsa.StevesCafe.com> from "Steve Passe" at Sep 7, 97 02:37:40 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Steve Passe wrote: > > Hi, > > I propose that we create an alternate file to smp.h named up.h. > > code could then look like: > > #ifdef SMP > #include > #else > #include > #endif > > I could then move stuff like: > > #ifndef SMP > #define THIS_LOCK /* nop this out for UP */ > #define THAT_FEATURE /* nop this out for UP */ > #endif > > from individual files and gather them all up into up.h This should cause less > disturbance to the source tree than the current methods. > alternative: replace smp.h and up.h with xxx.h all files include xxx.h in place of smp.h or up.h in xxx.h have a single conditional #ifndef SMP #define "up stuff" #else #define "smp stuff" #endif SMP is defined or not by /usr/sbin/config this concentrates all the conditionals into a single place. the source files only have one include file to deal with. just my two cents. jmb ps replace xxx.h with something reasonable.... From owner-freebsd-smp Sun Sep 7 22:38:12 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id WAA08187 for smp-outgoing; Sun, 7 Sep 1997 22:38:12 -0700 (PDT) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.129.19]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA08081; Sun, 7 Sep 1997 22:36:05 -0700 (PDT) Received: from critter.freebsd.dk (localhost.cybercity.dk [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.7) with ESMTP id HAA11171; Mon, 8 Sep 1997 07:34:21 +0200 (CEST) To: Steve Passe cc: smp@freebsd.org, cvs-committers@freebsd.org, cvs-all@freebsd.org, cvs-CVSROOT@freebsd.org Subject: Re: machine/up.h In-reply-to: Your message of "Sun, 07 Sep 1997 14:37:40 MDT." <199709072037.OAA27121@Ilsa.StevesCafe.com> Date: Mon, 08 Sep 1997 07:34:19 +0200 Message-ID: <11169.873696859@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Why not simply have the code do a #include and in that file have #ifdef SMP #define THIS_STUFF (lkjsdfgskdj) #define THAT_STUFF (lkjsdfgskdj) #else /* !SMP */ #define THIS_STUFF /* Nothing */ #define THAT_STUFF /* Nothing */ #endif In message <199709072037.OAA27121@Ilsa.StevesCafe.com>, Steve Passe writes: >Hi, > >I propose that we create an alternate file to smp.h named up.h. > >code could then look like: > >#ifdef SMP >#include >#else >#include >#endif > >I could then move stuff like: > >#ifndef SMP >#define THIS_LOCK /* nop this out for UP */ >#define THAT_FEATURE /* nop this out for UP */ >#endif > >from individual files and gather them all up into up.h This should cause less >disturbance to the source tree than the current methods. > >Any objections? > > >-- >Steve Passe | powered by >smp@csn.net | Symmetric MultiProcessor FreeBSD > > -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." From owner-freebsd-smp Mon Sep 8 10:41:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id KAA15063 for smp-outgoing; Mon, 8 Sep 1997 10:41:01 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id KAA15040; Mon, 8 Sep 1997 10:40:56 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id LAA00992; Mon, 8 Sep 1997 11:40:49 -0600 (MDT) Message-Id: <199709081740.LAA00992@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: smp@freebsd.org cc: current@freebsd.org Subject: does it fix things? Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 08 Sep 1997 11:40:48 -0600 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, So far I only have 1 report (it fixes the ip_fw lkm panic) on the fix John committed saturday. It fixes both John's and my test case, how about the rest of you? I had so many reports the last 3 weeks that I am just filing that pile, believing most if not all are related to the fix. If after applying the most recent code you still have problems re-post them. PS: I will be out of the office most of the week, so I wont be doing much on SMP... -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Mon Sep 8 11:46:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id LAA23058 for smp-outgoing; Mon, 8 Sep 1997 11:46:00 -0700 (PDT) Received: from atlantis.nconnect.net (root@atlantis.nconnect.net [207.227.50.2]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id LAA23027; Mon, 8 Sep 1997 11:45:50 -0700 (PDT) Received: from arabian (arabian.microxp.com [207.227.65.13]) by atlantis.nconnect.net (8.8.4/8.7.3) with SMTP id NAA05630; Mon, 8 Sep 1997 13:50:42 -0500 (CDT) Received: by localhost with Microsoft MAPI; Mon, 8 Sep 1997 13:52:10 -0500 Message-ID: <01BCBC5E.67384560.randyd@nconnect.net> From: Randy DuCharme Reply-To: "randyd@nconnect.net" To: "'Steve Passe'" , "smp@FreeBSD.ORG" , "'randy@microxp.com'" Cc: "current@FreeBSD.ORG" Subject: RE: does it fix things? Date: Mon, 8 Sep 1997 13:52:09 -0500 X-Mailer: Microsoft Internet E-mail/MAPI - 8.0.0.4211 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Monday, September 08, 1997 12:41 PM, Steve Passe [SMTP:smp@csn.net] wrote: > Hi, > > So far I only have 1 report (it fixes the ip_fw lkm panic) on the fix > John committed saturday. It fixes both John's and my test case, > how about the rest of you? > The only problem I had ( that I knew about ) was with the daemon screen saver causing the fatal trap 12. This *has* been fixed on my machine(s)! --- Randall D DuCharme Novell, Microsoft and UNIX Systems Engineer Networking Service & Support Computer Specialists BSD/OS Authorized Resellers & 414-253-9998 414-253-9919 (fax) BSDI Internet Success Partners From owner-freebsd-smp Mon Sep 8 12:44:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA29257 for smp-outgoing; Mon, 8 Sep 1997 12:44:18 -0700 (PDT) Received: from pluto.plutotech.com (ken@mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA29220; Mon, 8 Sep 1997 12:44:08 -0700 (PDT) Received: (from ken@localhost) by pluto.plutotech.com (8.8.5/8.8.5) id NAA21832; Mon, 8 Sep 1997 13:43:55 -0600 (MDT) From: Kenneth Merry Message-Id: <199709081943.NAA21832@pluto.plutotech.com> Subject: Re: does it fix things? In-Reply-To: <199709081740.LAA00992@Ilsa.StevesCafe.com> from Steve Passe at "Sep 8, 97 11:40:48 am" To: smp@csn.net (Steve Passe) Date: Mon, 8 Sep 1997 13:43:55 -0600 (MDT) Cc: smp@freebsd.org, current@freebsd.org X-Mailer: ELM [version 2.4ME+ PL28s (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Steve Passe wrote... > So far I only have 1 report (it fixes the ip_fw lkm panic) on the fix > John committed saturday. It fixes both John's and my test case, > how about the rest of you? > > I had so many reports the last 3 weeks that I am just filing that pile, > believing most if not all are related to the fix. If after applying the > most recent code you still have problems re-post them. > > PS: > I will be out of the office most of the week, so I wont be doing > much on SMP... It seems to fix things for me. My box made it through a complete make world, no problem. One caveat, though -- I'm not using any lkm's. Ken -- Kenneth Merry ken@plutotech.com From owner-freebsd-smp Mon Sep 8 13:04:59 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA00525 for smp-outgoing; Mon, 8 Sep 1997 13:04:59 -0700 (PDT) Received: from ic.net (qmailr@srv2b.ic.net [152.160.72.21]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id NAA00514 for ; Mon, 8 Sep 1997 13:04:53 -0700 (PDT) Received: (qmail 29873 invoked from network); 8 Sep 1997 20:03:59 -0000 Received: from unknown (HELO lurch.rickl.org) (152.160.108.56) by unknown with SMTP; 8 Sep 1997 20:03:59 -0000 Received: from lurch.rickl.org (rickl@localhost [127.0.0.1]) by lurch.rickl.org (8.8.7/8.7.3) with ESMTP id QAA02074; Mon, 8 Sep 1997 16:03:10 -0400 (EDT) Message-Id: <199709082003.QAA02074@lurch.rickl.org> Date: Mon, 8 Sep 1997 16:03:08 -0400 (EDT) From: rickl@ic.net Reply-To: rickl@ic.net Subject: Re: does it fix things? To: smp@csn.net cc: smp@freebsd.org, current@freebsd.org In-Reply-To: <199709081740.LAA00992@Ilsa.StevesCafe.com> MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On 8 Sep, Steve Passe was accused of saying: > Hi, > > So far I only have 1 report (it fixes the ip_fw lkm panic) on the fix > John committed saturday. It fixes both John's and my test case, > how about the rest of you? > > rest of message deleted...... Steve, All is better here as well. On a somewhat unrelated issue, I noticed that the serial link in interrupted during audio processing (e.g. when FvwmAudio is trying to play something). Is this a further manifestation of the IDE disk access/sio problems I noticed last week?? Is this only a problem on my system?? Has anyone else noticed it?? Rick From owner-freebsd-smp Mon Sep 8 22:46:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id WAA04944 for smp-outgoing; Mon, 8 Sep 1997 22:46:06 -0700 (PDT) Received: from atlantis.nconnect.net (root@atlantis.nconnect.net [207.227.50.2]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA04934 for ; Mon, 8 Sep 1997 22:46:00 -0700 (PDT) Received: from nconnect.net (randyd@dial181.nconnect.net [207.227.50.181]) by atlantis.nconnect.net (8.8.4/8.7.3) with ESMTP id AAA05353 for ; Tue, 9 Sep 1997 00:50:51 -0500 (CDT) Message-ID: <3414E1F2.9C48CCCE@nconnect.net> Date: Tue, 09 Sep 1997 00:43:14 -0500 From: Randy DuCharme Organization: Astrolab Development X-Mailer: Mozilla 4.02b7 [en] (X11; I; FreeBSD 3.0-CURRENT i386) MIME-Version: 1.0 To: smp@freebsd.org Subject: silo overflows gone! Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Greetings, Just a point of information (FWIW) .... The latest kernel changes seems to have completely eliminated silo overflow messages on my machine. I've been accustomed to seeing a few here and there up to several dozen per hour. I've just downloaded over 40 Megabytes worth of junk at 4.8 - 5.0 K/sec and a port speed of 115200 without a single message! Or did you guys just turn off the messages?? 8-) -- Randall D DuCharme Systems Engineer Novell, Microsoft, and UNIX Networking Support Computer Specialists BSDI Internet Success Partners 414-253-9998 414-253-9919 (fax) BSD/OS Authorized Resellers From owner-freebsd-smp Tue Sep 9 18:25:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id SAA22561 for smp-outgoing; Tue, 9 Sep 1997 18:25:22 -0700 (PDT) Received: from abby.skypoint.net (abby.skypoint.net [199.86.32.252]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id SAA22554 for ; Tue, 9 Sep 1997 18:25:17 -0700 (PDT) Received: (from uucp@localhost) by abby.skypoint.net (8.8.5/alexis 2.7) with UUCP id UAA25872; Tue, 9 Sep 1997 20:25:09 -0500 (CDT) Received: (from bruce@localhost) by zuhause.mn.org (8.8.7/8.8.5) id UAA04902; Tue, 9 Sep 1997 20:24:51 -0500 (CDT) Date: Tue, 9 Sep 1997 20:24:51 -0500 (CDT) Message-Id: <199709100124.UAA04902@zuhause.mn.org> From: Bruce Albrecht To: Randy DuCharme Cc: smp@freebsd.org Subject: Re: silo overflows gone! In-Reply-To: <3414E1F2.9C48CCCE@nconnect.net> References: <3414E1F2.9C48CCCE@nconnect.net> X-Mailer: VM 6.30 under 19.15p2 XEmacs Lucid Mime-Version: 1.0 (generated by tm-edit 7.106) Content-Type: text/plain; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Randy DuCharme writes: > Greetings, > > Just a point of information (FWIW) .... > > The latest kernel changes seems to have completely eliminated silo > overflow > messages on my machine. I've been accustomed to seeing a few here and > there > up to several dozen per hour. I've just downloaded over 40 Megabytes > worth of > junk at 4.8 - 5.0 K/sec and a port speed of 115200 without a single > message! I rebuilt my kernel Saturday, and I've seen a couple of silo overflows at 57600, but it went from ~40 on Saturday to 2 on Sunday, although I don't think I transferred as much stuff on Sunday. From owner-freebsd-smp Thu Sep 11 14:49:59 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id OAA23624 for smp-outgoing; Thu, 11 Sep 1997 14:49:59 -0700 (PDT) Received: from cod.nosc.mil (root@cod.nosc.mil [128.49.4.5]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id OAA23618 for ; Thu, 11 Sep 1997 14:49:56 -0700 (PDT) Received: from [128.49.16.48] (aegis.nosc.mil [128.49.16.48]) by cod.nosc.mil (8.8.7/8.8.7) with ESMTP id OAA14552 for ; Thu, 11 Sep 1997 14:49:51 -0700 (PDT) X-Sender: gshaffer@cod.nosc.mil Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 11 Sep 1997 14:47:37 -0700 To: smp@freebsd.org From: Greg Shaffer Subject: Re: Make and SMP - what can be done ? Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I sent this to current the other day and got no response, so I thought I would send it to smp and see what happens. >X-Sender: gshaffer@cod.nosc.mil >Mime-Version: 1.0 >Date: Tue, 9 Sep 1997 12:09:19 -0700 >To: current@FreeBSD.ORG >From: Greg Shaffer >Subject: Re: Make and SMP - what can be done ? >Cc: gshaffer@cod.nosc.mil >Sender: owner-freebsd-current@FreeBSD.ORG >X-Loop: FreeBSD.org >Precedence: bulk > >Around the middle of August there was some discussion on how to make 'make >world' operate safely in a parallel mode. There were even some patches to >make, *.mk and Makefiles to support this. I applied these patches to my >system and was very impressed with the performance improvement. Did these >patches ever get roles into the source tree, it would be a shame to loss >them. > >Thanks >Greg Shaffer > From owner-freebsd-smp Fri Sep 12 17:01:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA19560 for smp-outgoing; Fri, 12 Sep 1997 17:01:29 -0700 (PDT) Received: from gw.teledigit.se ([193.14.252.3]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id RAA19555 for ; Fri, 12 Sep 1997 17:01:25 -0700 (PDT) Received: (from smap@localhost) by gw.teledigit.se (8.7.5/8.7.3) id BAA26185 for ; Sat, 13 Sep 1997 01:47:44 +0200 (MET DST) Message-Id: <199709122347.BAA26185@gw.teledigit.se> X-Authentication-Warning: gw.teledigit.se: smap set sender to using -f Received: from unknown(192.168.16.22) by gw.teledigit.se via smap (V1.3) id sma026183; Sat Sep 13 01:47:36 1997 Received: by mal-cr.teledigit.se (NX5.67e/NX3.0X) id AA16299; Sat, 13 Sep 97 02:08:41 +0200 Content-Type: text/plain Mime-Version: 1.0 (NeXT Mail 3.3 v118.2) Received: by NeXT.Mailer (1.118.2) From: Robert Lindgren Date: Sat, 13 Sep 97 02:08:39 +0200 To: freebsd-smp@freebsd.org Subject: How to fire up the second CPU? Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I have a dual pp 200 with 256 meg ram on a Intel PR440FX and I don't get the second cpu up. Anyone have a suggestion? I have enable the SMP-GENERIC kernel and I'm running on FreeBSD-3.0-970907-SNAP. Please answer me if you can Robert From owner-freebsd-smp Sat Sep 13 11:22:51 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id LAA18731 for smp-outgoing; Sat, 13 Sep 1997 11:22:51 -0700 (PDT) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id LAA18726 for ; Sat, 13 Sep 1997 11:22:47 -0700 (PDT) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by mail.cdsnet.net (8.8.6/8.8.6) with SMTP id LAA27880 for ; Sat, 13 Sep 1997 11:22:42 -0700 (PDT) Date: Sat, 13 Sep 1997 11:22:41 -0700 (PDT) From: Jaye Mathisen To: smp@freebsd.org Subject: mildly amusing SMP problem... Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I've supped as of 9/11/97, and built a new kernel. Right after it prints SMP: CPU#1 luanched console outout just stops. The system does continue booting, and comes up, but can't see jack. Dropping back to an earlier kernel works fine. The other problem is that after a make world and reboot, whenever I login with ssh, my screen is always 1 character behind (in xterm). ie, I'm at the # prompt. I type an l, nothing happens. type the s, the l appears. Hit return, the s appears. Hit return again, I get the output. I'm hoping it's just some weird ssh issue, and am recompiling to make sure. But the console output is just bonkers. Nope, rebuilding ssh and restring it didn't do it. Maybe it's an XFree86 problem. Using telnet seems to work OK. From owner-freebsd-smp Sat Sep 13 11:51:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id LAA20354 for smp-outgoing; Sat, 13 Sep 1997 11:51:25 -0700 (PDT) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id LAA20343 for ; Sat, 13 Sep 1997 11:51:20 -0700 (PDT) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by mail.cdsnet.net (8.8.6/8.8.6) with SMTP id LAA02932 for ; Sat, 13 Sep 1997 11:51:17 -0700 (PDT) Date: Sat, 13 Sep 1997 11:51:16 -0700 (PDT) From: Jaye Mathisen To: smp@freebsd.org Subject: Definitely seems SMP related. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Rebuilding from the same sources with SMP and APIO commented out, and the kernel works just fine. All console output appears, ssh/xterm combo works fine. Reboot with SMP kernel, and it wigs out again. One other thing I noticed, is that when booting the SMP kernel, the reboot command fails. It works in that my IP connection disappears, but the machine never restarts. In UNI mode, it works fine. From owner-freebsd-smp Sat Sep 13 17:52:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA09736 for smp-outgoing; Sat, 13 Sep 1997 17:52:57 -0700 (PDT) Received: from milo.cfw.com (milo.cfw.com [205.219.240.6]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id RAA09731 for ; Sat, 13 Sep 1997 17:52:54 -0700 (PDT) Message-Id: <199709140052.RAA09731@hub.freebsd.org> Received: (qmail 26904 invoked from network); 14 Sep 1997 00:54:18 -0000 Received: from ras12wb8.cfw.com (HELO pauls2) (208.217.184.209) by milo.cfw.com with SMTP; 14 Sep 1997 00:54:18 -0000 From: "Paul Missman" To: Subject: SMP in FreeBSD 3.x.x Date: Sat, 13 Sep 1997 20:56:29 -0400 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 4.71.1008.3 X-MimeOle: Produced By Microsoft MimeOLE Engine V4.71.1008.3 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I see that the SMP source was merged with the 3.x.x development kernal. Does this mean that 3.x.x, when released, will seamlessly support 1 - N processors? Also, what do you imagine maximum N will be at 3.x.x release time? I believe the original Intel spec was up to 4 processors, but I see a lot of 8 processor servers coming out lately. Thanks, Paul Missman From owner-freebsd-smp Sat Sep 13 21:24:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id VAA19993 for smp-outgoing; Sat, 13 Sep 1997 21:24:18 -0700 (PDT) Received: from usr08.primenet.com (tlambert@usr08.primenet.com [206.165.6.208]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id VAA19986 for ; Sat, 13 Sep 1997 21:24:11 -0700 (PDT) Received: (from tlambert@localhost) by usr08.primenet.com (8.8.5/8.8.5) id VAA07740; Sat, 13 Sep 1997 21:24:08 -0700 (MST) From: Terry Lambert Message-Id: <199709140424.VAA07740@usr08.primenet.com> Subject: Re: SMP in FreeBSD 3.x.x To: missmanp@milo.cfw.com (Paul Missman) Date: Sun, 14 Sep 1997 04:24:08 +0000 (GMT) Cc: freebsd-smp@freebsd.org In-Reply-To: <199709140052.RAA09731@hub.freebsd.org> from "Paul Missman" at Sep 13, 97 08:56:29 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > I see that the SMP source was merged with the 3.x.x development kernal. > Does this mean that 3.x.x, when released, will seamlessly support 1 - N > processors? Also, what do you imagine maximum N will be at 3.x.x release > time? I believe the original Intel spec was up to 4 processors, but I see a > lot of 8 processor servers coming out lately. A system can have up to 32 APICs. It is the APIC IDs that limit the number of processors. Given that a system will have an I/O APIC, the max is 31 (except for I2O systems, where the i960 also has an APIC, or systems with multiple bus controllers -- each with an APIC). So ~30. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-smp Sat Sep 13 23:47:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id XAA25459 for smp-outgoing; Sat, 13 Sep 1997 23:47:03 -0700 (PDT) Received: from spinner.dialix.com.au (spinner.dialix.com.au [202.12.86.3]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id XAA25440 for ; Sat, 13 Sep 1997 23:46:55 -0700 (PDT) Received: from spinner.dialix.com.au (localhost.dialix.com.au [127.0.0.1]) by spinner.dialix.com.au with ESMTP id OAA22630; Sun, 14 Sep 1997 14:46:25 +0800 (WST) Message-Id: <199709140646.OAA22630@spinner.dialix.com.au> X-Mailer: exmh version 2.0zeta 7/24/97 To: Terry Lambert cc: missmanp@milo.cfw.com (Paul Missman), freebsd-smp@freebsd.org Subject: Re: SMP in FreeBSD 3.x.x In-reply-to: Your message of "Sun, 14 Sep 1997 04:24:08 GMT." <199709140424.VAA07740@usr08.primenet.com> Date: Sun, 14 Sep 1997 14:46:24 +0800 From: Peter Wemm Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Terry Lambert wrote: > > I see that the SMP source was merged with the 3.x.x development kernal. > > Does this mean that 3.x.x, when released, will seamlessly support 1 - N > > processors? Also, what do you imagine maximum N will be at 3.x.x release > > time? I believe the original Intel spec was up to 4 processors, but I see a > > lot of 8 processor servers coming out lately. > > A system can have up to 32 APICs. It is the APIC IDs that limit > the number of processors. > > Given that a system will have an I/O APIC, the max is 31 (except > for I2O systems, where the i960 also has an APIC, or systems with > multiple bus controllers -- each with an APIC). > > So ~30. No, the APIC id's go from 0 to 15, and if memory serves, "15" is used as an 'all processors' value in some places.. The limit is either 15 or 16 APIC device id's depending on that. Since there should be at least one IO APIC (or there's not much point doing SMP), that makes the limit either 14 or 15 cpus. This is for a conventional APIC bus structure of course. :-) The MPSPEC (or was it the pentium family hardware docs?) talks about APIC bridges and wild things like that which could (I think) take the device limit up higher (60 devices seems to ring a bell), but it's the least of our worries at the moment. :-) The older 82489DX i486 APIC had an 8-bit device ID and a 4-wire APIC bus (the p5/p6 family have a 2-wire bus with 4-bit id's). A non-intel system such as OpenPIC would have different constraints, assuming the hardware actually exists. > Terry Lambert > terry@lambert.org > --- > Any opinions in this posting are my own and not those of my present > or previous employers. > Cheers, -Peter