From owner-freebsd-smp Sun Apr 13 00:41:42 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id AAA22413 for smp-outgoing; Sun, 13 Apr 1997 00:41:42 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA22401 for ; Sun, 13 Apr 1997 00:41:39 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id BAA03311; Sun, 13 Apr 1997 01:41:32 -0600 (MDT) Message-Id: <199704130741.BAA03311@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Bruce Albrecht cc: freebsd-smp@freebsd.org Subject: Re: Tyan ATX 1668 In-reply-to: Your message of "Sat, 12 Apr 1997 13:28:57 CDT." <9704121828.AA26206@gf006e0.seag.fingerhut.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 13 Apr 1997 01:41:32 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > I've got one of these on order, and I understand that there's a > ... > it matters which BIOS (AMI/Award) one is using. If it's BIOS, what's > the change that Tyan will have the BIOS manufacturers fix it if we all > complain? about 0.1% it can be worked around, its just a pain in the neck... --- > Also, this will be my first FreeBSD system (currently running > NetBSD-Amiga), and I was wondering if I should start out running > FreeBSD 2.2.1, and start looking at FreeBSD-SMP after I've gotten my > basic setup done, or take the plunge and go start to FreeBSD-3.0+SMP > kernel? Would I run 3.0 software + SMP kernel, or is there a complete > SMP tree? I suggest starting with the 3.0-970209-SNAP + SMP kernel tree. this is the point just the 2 were last merged, as is reasomnably stable. I have this combination running on both a dual P5 and dual P6. for details on installing see: http://www.freebsd.org/~fsmp/SMP/SMP.html -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sun Apr 13 06:19:07 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA03905 for smp-outgoing; Sun, 13 Apr 1997 06:19:07 -0700 (PDT) Received: from charon.finall.com (charon.finall.com [206.246.160.131]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA03898 for ; Sun, 13 Apr 1997 06:19:03 -0700 (PDT) Received: from exchange.finall.com (exchange.finall.com [206.246.160.132]) by charon.finall.com (8.8.5/8.6.12) with SMTP id JAA13914 for ; Sun, 13 Apr 1997 09:19:02 -0400 (EDT) Received: by exchange.finall.com with Microsoft Exchange (IMC 4.0.837.3) id <01BC47EB.BCE19BC0@exchange.finall.com>; Sun, 13 Apr 1997 09:19:07 -0400 Message-ID: From: "Jung, Michael" To: "'Bruce.Albrecht@seag.fingerhut.com'" , "'terry@lambert.org'" , "'vanmaren@fast.cs.utah.edu'" Cc: "'freebsd-smp@freebsd.org'" Subject: RE: Tyan ATX 1668 Date: Sun, 13 Apr 1997 09:19:04 -0400 X-Mailer: Microsoft Exchange Server Internet Mail Connector Version 4.0.837.3 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk After escalating a problem with a 3940W & NT 4.0, Adaptec basically duplicated the problem, then Tyan, then Award. We were shipped a AMI flash for the Tomcat III and everything works great. Please note that the problem we had was that any drives that were on the second channel were not available to the OS. Unfortunately the machine is in production and I can't do any testing with FreeBSD at this point. I just wanted to let everyone know that AMI BIOS does exist. As a last note, we received the new flash last week 4/2/97. --mikej Michael Jung mikej@finall.com >---------- >From: vanmaren@fast.cs.utah.edu[SMTP:vanmaren@fast.cs.utah.edu] >Sent: Sunday, April 13, 1997 1:40 AM >To: Bruce.Albrecht@seag.fingerhut.com; terry@lambert.org >Cc: freebsd-smp@freebsd.org >Subject: Re: Tyan ATX 1668 > >>> I've got one of these on order, and I understand that there's a >>> problem with the 5th PCI slot, which is not currently a problem for me >>> since I'm currently planning to use 3 slots. However, I was wondering >>> if this problem is a hardware problem or a BIOS problem, and whether >>> it matters which BIOS (AMI/Award) one is using. If it's BIOS, what's >>> the change that Tyan will have the BIOS manufacturers fix it if we all >>> complain? > >Only the Award BIOS is available from Tyan (If you can find the >AMI, let me know; I'd really like to try it!). > > > From owner-freebsd-smp Mon Apr 14 11:59:18 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA06459 for smp-outgoing; Mon, 14 Apr 1997 11:59:18 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id LAA06452 for ; Mon, 14 Apr 1997 11:59:15 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA19007; Mon, 14 Apr 1997 11:37:44 -0700 From: Terry Lambert Message-Id: <199704141837.LAA19007@phaeton.artisoft.com> Subject: Re: Tyan ATX 1668 To: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Date: Mon, 14 Apr 1997 11:37:43 -0700 (MST) Cc: Bruce.Albrecht@seag.fingerhut.com, terry@lambert.org, freebsd-smp@freebsd.org In-Reply-To: <199704130540.XAA15914@fast.cs.utah.edu> from "Kevin Van Maren" at Apr 12, 97 11:40:28 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > >The BIOS problem is that the mptable does not refect the correct number > >of PCI busses on the motherboard (it says there is one and there are 2), > >and so the second bus's APIC is incorrectly configured to an ISA > >interrupt (you can get more information on this by searching the SMP > >list archives on www.freebsd.org). > > No. They Tyan S1662/S1668 has only 1 (one) PCI bus. All 5 slots > are on the same PCI bus. Of course, there are two PCI busses if > you use an Adaptec 3940 or other device with a PCI-PCI bridge on it. > But the particular problem referred to was the missing entry for > the 5th slot. I stand corrected. You know, of course, that there isn't a 5th slot, right? It's a second fourth slot. It's probably more correct to claim that they don't have an entry for a fifth device... Regards, 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 Mon Apr 14 12:20:01 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA07872 for smp-outgoing; Mon, 14 Apr 1997 12:20:01 -0700 (PDT) Received: from cs.utah.edu (cs.utah.edu [128.110.4.21]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA07834 for ; Mon, 14 Apr 1997 12:19:58 -0700 (PDT) Received: from fast.cs.utah.edu by cs.utah.edu (8.8.4/utah-2.21-cs) id NAA17709; Mon, 14 Apr 1997 13:18:37 -0600 (MDT) Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id NAA13963; Mon, 14 Apr 1997 13:18:36 -0600 Date: Mon, 14 Apr 1997 13:18:36 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Message-Id: <199704141918.NAA13963@fast.cs.utah.edu> To: terry@lambert.org Subject: Re: Tyan ATX 1668 Cc: freebsd-smp@freebsd.org Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >I stand corrected. > >You know, of course, that there isn't a 5th slot, right? It's a >second fourth slot. Huh? There *are* *5* slots. Look at http://www.tyan.com/html/s166{2,8}.html for a picture. >It's probably more correct to claim that they don't have an entry >for a fifth device... > Here is `part' of a dmesg output (stripped out disks, etc) for a S1662 motherboard. Only one processor right now, but I'm using this one since it has 5 PCI cards in it (plus ISA video). chip0 rev 2 on pci0:0 chip1 rev 1 on pci0:7:0 chip2 rev 0 on pci0:7:1 chip3 rev 1 on pci0:10 ahc0 rev 0 int a irq 5 on pci0:11 ahc0: aic7880 Wide Channel, SCSI Id=7, 16 SCBs de0 rev 17 int a irq 9 on pci0:12 de0: SMC 9332 DC21140 [10-100Mb/s] pass 1.1 de0: address 00:00:c0:50:0c:9f de0: enabling 100baseTX port ahc1 rev 0 int a irq 10 on pci0:13 ahc1: aic7870 Wide Channel, SCSI Id=7, 16 SCBs de1 rev 18 int a irq 11 on pci0:14 de1: SMC 9332 DC21140 [10-100Mb/s] pass 1.2 de1: address 00:00:c0:43:65:e4 de1: enabling 100baseTX port Probing for devices on PCI bus 1: de2 rev 35 int a irq 11 on pci1:4 de2: DC21040 [10Mb/s] pass 2.3 de2: address 00:c0:95:f4:04:ba de2: enabling BNC/AUI port de3 rev 35 int a irq 10 on pci1:5 de3: (null)DC21040 [10Mb/s] pass 2.3 de3: address 00:c0:95:f4:04:bb de3: enabling BNC/AUI port PCI Bus #0 has 5 slots: 10: Pci bridge 4: Network (de2,irq 11) 5: Network (de3,irq 11, says 10) 11: Mass Storage (ahc0,irq 5) 12: Network (de0,irq 9) 13: Mass Storage (de1,irq 10) 14: Network (ahc1,irq 11) This is with FreeBSD 2.1.7.1; de2/de3 is a Zynx dual 10Mb ethernet adapter. Kevin From owner-freebsd-smp Mon Apr 14 14:12:08 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA14586 for smp-outgoing; Mon, 14 Apr 1997 14:12:08 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id OAA14563 for ; Mon, 14 Apr 1997 14:12:04 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id NAA19171; Mon, 14 Apr 1997 13:19:16 -0700 From: Terry Lambert Message-Id: <199704142019.NAA19171@phaeton.artisoft.com> Subject: Re: Tyan ATX 1668 To: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Date: Mon, 14 Apr 1997 13:19:16 -0700 (MST) Cc: terry@lambert.org, freebsd-smp@freebsd.org In-Reply-To: <199704141918.NAA13963@fast.cs.utah.edu> from "Kevin Van Maren" at Apr 14, 97 01:18:36 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > >I stand corrected. > > > >You know, of course, that there isn't a 5th slot, right? It's a > >second fourth slot. > > Huh? > > There *are* *5* slots. > Look at http://www.tyan.com/html/s166{2,8}.html for a picture. The 4th and 5th slot are electrically equivalent. There are only 4 PCI interrupts, and a direct cascade mand 5 & 1 have the same ordering and a non-direct cascade meand that 4 & 5 have the same ordering; in the case of that particular motherboard, $ & 5 are electrically equivalent in all respects except bus arbitration (and that's irrelevant at the slot level, at least for PCI). > >It's probably more correct to claim that they don't have an entry > >for a fifth device... > > > > Here is `part' of a dmesg output (stripped out disks, etc) for > a S1662 motherboard. Only one processor right now, but I'm using > this one since it has 5 PCI cards in it (plus ISA video). [ ... ] > PCI Bus #0 has 5 slots: > 10: Pci bridge > 4: Network (de2,irq 11) > 5: Network (de3,irq 11, says 10) > 11: Mass Storage (ahc0,irq 5) > 12: Network (de0,irq 9) > 13: Mass Storage (de1,irq 10) > 14: Network (ahc1,irq 11) > > This is with FreeBSD 2.1.7.1; de2/de3 is a Zynx dual 10Mb ethernet > adapter. This is with the updated BIOS, right? The "5 slots" are a device iteration count, not a true slot count (as above). Regards, 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 Mon Apr 14 14:23:32 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id OAA15174 for smp-outgoing; Mon, 14 Apr 1997 14:23:32 -0700 (PDT) Received: from agora.rdrop.com (root@agora.rdrop.com [199.2.210.241]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id OAA15165 for ; Mon, 14 Apr 1997 14:23:29 -0700 (PDT) Received: from jingoro.prevmed.sunysb.edu by agora.rdrop.com with smtp (Smail3.1.29.1 #17) id m0wGtDi-00091OC; Mon, 14 Apr 97 14:23 PDT Received: (from cfanning@localhost) by jingoro.prevmed.sunysb.edu (8.8.3/8.6.12) id RAA29270 for freebsd-smp@freebsd.org; Mon, 14 Apr 1997 17:13:52 -0400 (EDT) From: Chris Fanning Message-Id: <199704142113.RAA29270@jingoro.prevmed.sunysb.edu> Subject: Re: Tyan ATX 1668 To: freebsd-smp@freebsd.org Date: Mon, 14 Apr 1997 17:13:52 -0400 (EDT) Content-Type: text Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk * warning: I haven't done this myself If you check Tyan's FTP site there is a directory which has an AMI BIOS for the board. I haven't flashed this myself so I have no idea if it works at all. Someone said that the AMI BIOS Tyan supplied them with worked in another board after the Award failed them. Might be worth checking out. There's also an "eval" directory for what looks like the next major revision of the Award BIOS. It's pretty safe to say that this isn't supported at all but it would be interesting to see if this fixed a few problems being reported here. Chris From owner-freebsd-smp Mon Apr 14 16:10:50 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id QAA22248 for smp-outgoing; Mon, 14 Apr 1997 16:10:50 -0700 (PDT) Received: from hobbes.saturn-tech.com (drussell@[207.229.19.5]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA22231 for ; Mon, 14 Apr 1997 16:10:45 -0700 (PDT) Received: from localhost (drussell@localhost) by hobbes.saturn-tech.com (8.8.4/8.8.2) with SMTP id RAA09205; Mon, 14 Apr 1997 17:07:20 -0600 (MDT) Date: Mon, 14 Apr 1997 17:07:20 -0600 (MDT) From: Doug Russell To: Kevin Van Maren cc: terry@lambert.org, freebsd-smp@freebsd.org Subject: Re: Tyan ATX 1668 In-Reply-To: <199704141918.NAA13963@fast.cs.utah.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > Huh? > > There *are* *5* slots. > Look at http://www.tyan.com/html/s166{2,8}.html for a picture. Physically there are 5 slots, but electrically slot 4 and 5 are the same. The PCI slots have 4 interrupts, each slot being rotated one further down the interrupt chain. Slot 5 is just another slot 4, which, does work just fine under virtually any circumstance if you use one of them for a video card or something else that doesn't need an interrupt line. Later...... From owner-freebsd-smp Mon Apr 14 17:25:46 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id RAA27108 for smp-outgoing; Mon, 14 Apr 1997 17:25:46 -0700 (PDT) Received: from uhf.wdc.net (uhf.wdc.net [198.147.74.44]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA27102 for ; Mon, 14 Apr 1997 17:25:42 -0700 (PDT) Received: from localhost (bad@localhost) by uhf.wdc.net (8.8.5/8.6.12) with SMTP id UAA02341 for ; Mon, 14 Apr 1997 20:26:13 -0400 (EDT) Date: Mon, 14 Apr 1997 20:26:12 -0400 (EDT) From: Bernie Doehner To: smp@freebsd.org Subject: Any good deals on PPro Motherboards? Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi Gang: Do any of you know of any good PPro Motherboard deals for MB's that are on the "good" list? Specificaly, I am looking for a board that has SDRAM sockets. Thanks. Bernie From owner-freebsd-smp Mon Apr 14 21:26:20 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA13128 for smp-outgoing; Mon, 14 Apr 1997 21:26:20 -0700 (PDT) Received: from INET-02-IMC.microsoft.com (mail2.microsoft.com [131.107.3.42]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA13122 for ; Mon, 14 Apr 1997 21:26:18 -0700 (PDT) Received: by mail2.microsoft.com with Internet Mail Service (5.0.1458.8) id <29WMYYB7>; Mon, 14 Apr 1997 21:21:41 -0700 Message-ID: <7D06B4AA8B39D011A64900805F682CDA01001D72@RED-09-MSG.dns.microsoft.com> From: Arlie Davis To: "'Doug Russell'" , Kevin Van Maren Cc: terry@lambert.org, freebsd-smp@freebsd.org Subject: RE: Tyan ATX 1668 Date: Mon, 14 Apr 1997 17:09:05 -0700 X-Priority: 3 X-Mailer: Internet Mail Service (5.0.1458.8) Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Or if the drivers correctly support INT / IRQ sharing, the way God intended PCI drivers to work. My copy of the PCI 2.x spec says that INTs should always always always be shareable. The whole idea of having four INTs is to _reduce_ but not _eliminate_ INT channel sharing. For example, a downstream PCI bus (behind a PCI-to-PCI bridge) could have four PCI slots, each with four INTs. (The INTs would be shared in the usual round-robin fashion.) The INTs from that downstream bus are propagated upward to the main PCI bus onto the four INTs of the hosting PCI slot. This is why the Adaptec 3940 is so mind-blowing at first. But the idea is that you still _have to support_ INT sharing to fully support PCI. So, having five PCI slots is completely "legitimate". You should be able to put a PCI card in both slot 4 and 5 which both use INT A (or any common PCI INT) and it _should_ work just fine. It doesn't, of course, under most Intel-architecture OSes. -- arlie > -----Original Message----- > From: Doug Russell [SMTP:drussell@saturn-tech.com] > Sent: Monday, April 14, 1997 4:07 PM > To: Kevin Van Maren > Cc: terry@lambert.org; freebsd-smp@freebsd.org > Subject: Re: Tyan ATX 1668 > > > > Huh? > > > > There *are* *5* slots. > > Look at http://www.tyan.com/html/s166{2,8}.html for a picture. > > Physically there are 5 slots, but electrically slot 4 and 5 are the > same. > The PCI slots have 4 interrupts, each slot being rotated one further > down > the interrupt chain. Slot 5 is just another slot 4, which, does work > just > fine under virtually any circumstance if you use one of them for a > video > card or something else that doesn't need an interrupt line. > > Later...... > > From owner-freebsd-smp Tue Apr 15 10:21:05 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA22399 for smp-outgoing; Tue, 15 Apr 1997 10:21:05 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id KAA22390 for ; Tue, 15 Apr 1997 10:21:00 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id JAA22513; Tue, 15 Apr 1997 09:58:22 -0700 From: Terry Lambert Message-Id: <199704151658.JAA22513@phaeton.artisoft.com> Subject: Re: Tyan ATX 1668 To: arlied@microsoft.com (Arlie Davis) Date: Tue, 15 Apr 1997 09:58:22 -0700 (MST) Cc: drussell@saturn-tech.com, vanmaren@fast.cs.utah.edu, terry@lambert.org, freebsd-smp@freebsd.org In-Reply-To: <7D06B4AA8B39D011A64900805F682CDA01001D72@RED-09-MSG.dns.microsoft.com> from "Arlie Davis" at Apr 14, 97 05:09:05 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > So, having five PCI slots is completely "legitimate". You should be > able to put a PCI card in both slot 4 and 5 which both use INT A (or any > common PCI INT) and it _should_ work just fine. It doesn't, of course, > under most Intel-architecture OSes. It's not legitimate if the MPtable does not account for five PCI devices, so one of the devices is misidentified as an ISA device. You can't share interrupts between PCI and ISA devices. 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 Tue Apr 15 12:50:36 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id MAA05933 for smp-outgoing; Tue, 15 Apr 1997 12:50:36 -0700 (PDT) Received: from INET-02-IMC.microsoft.com (mail2.microsoft.com [131.107.3.42]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA05927 for ; Tue, 15 Apr 1997 12:50:34 -0700 (PDT) Received: by mail2.microsoft.com with Internet Mail Service (5.0.1458.14) id <20030B6A>; Tue, 15 Apr 1997 12:49:37 -0700 Message-ID: <7D06B4AA8B39D011A64900805F682CDA01001D7E@RED-09-MSG.dns.microsoft.com> From: Arlie Davis To: freebsd-smp@freebsd.org Cc: vanmaren@fast.cs.utah.edu, drussell@saturn-tech.com Subject: Re: Tyan ATX 1668 Date: Tue, 15 Apr 1997 12:04:11 -0700 X-Priority: 3 X-Mailer: Internet Mail Service (5.0.1458.14) Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk -----Original Message----- From: Terry Lambert [SMTP:terry@lambert.org] Sent: Tuesday, April 15, 1997 9:58 AM To: Arlie Davis Cc: drussell@saturn-tech.com; vanmaren@fast.cs.utah.edu; terry@lambert.org; freebsd-smp@freebsd.org Subject: Re: Tyan ATX 1668 > So, having five PCI slots is completely "legitimate". You should be > able to put a PCI card in both slot 4 and 5 which both use INT A (or any > common PCI INT) and it _should_ work just fine. It doesn't, of course, > under most Intel-architecture OSes. It's not legitimate if the MPtable does not account for five PCI devices, so one of the devices is misidentified as an ISA device. [Arlie Davis] Of course. The implementation is just broken. You can't share interrupts between PCI and ISA devices. [Arlie Davis] Of course. Nothing that I wrote contradicts this -- I spoke only of the PCI spec. From owner-freebsd-smp Tue Apr 15 20:12:26 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id UAA01948 for smp-outgoing; Tue, 15 Apr 1997 20:12:26 -0700 (PDT) Received: from 586quick166.saturn-tech.com ([207.229.19.32]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA01934 for ; Tue, 15 Apr 1997 20:12:21 -0700 (PDT) Received: from localhost (drussell@localhost) by 586quick166.saturn-tech.com (8.8.4/8.8.4) with SMTP id VAA12535; Tue, 15 Apr 1997 21:11:37 -0600 (MDT) X-Authentication-Warning: 586quick166.saturn-tech.com: drussell owned process doing -bs Date: Tue, 15 Apr 1997 21:11:36 -0600 (MDT) From: Doug Russell To: Arlie Davis cc: Kevin Van Maren , terry@lambert.org, freebsd-smp@freebsd.org Subject: RE: Tyan ATX 1668 In-Reply-To: <7D06B4AA8B39D011A64900805F682CDA01001D72@RED-09-MSG.dns.microsoft.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Mon, 14 Apr 1997, Arlie Davis wrote: > My copy of the PCI 2.x spec says that INTs should always always always > be shareable. The whole idea of having four INTs is to _reduce_ but not ... > common PCI INT) and it _should_ work just fine. It doesn't, of course, > under most Intel-architecture OSes. Hehe.. Exactly. I know it is SUPPOSED to work, that doesn't mean it always DOES work. :) Later...... From owner-freebsd-smp Wed Apr 16 15:07:45 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA18102 for smp-outgoing; Wed, 16 Apr 1997 15:07:45 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA18091; Wed, 16 Apr 1997 15:07:40 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id QAA18500; Wed, 16 Apr 1997 16:07:29 -0600 (MDT) Message-Id: <199704162207.QAA18500@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Michael Searle cc: freebsd-hardware@FreeBSD.ORG, smp@FreeBSD.ORG Subject: Re: SMP question. In-reply-to: Your message of "Wed, 16 Apr 1997 20:14:07 -0000." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 16 Apr 1997 16:07:28 -0600 Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi, > ... > > there is some possibility that the P-90 (I assumme you mean P5-90) will > > be a problem. That vintage of P5 is from the days when the APIC section > > was somewhat unstable, and often required a "matched pair" of CPUs, ie 2 > > from the same stepping and rev level. > > Is this ever necessary for a 2 or 4 P6-150? it is generally believed that this is NOT a problem with any of the P6 chips, but I can't guarantee it. you might also consider trying to clock the P5-150s at 166mHz, the bus will then run @ 66mHz instead of 60mHz. I just purchased a dual P6 with 2 P6-166mHzx512k cache and am successfully running them at 200mHz. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Fri Apr 18 03:50:01 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id DAA23977 for smp-outgoing; Fri, 18 Apr 1997 03:50:01 -0700 (PDT) Received: from ceylon.informatik.uni-rostock.de (ceylon.informatik.uni-rostock.de [139.30.5.237]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA23951 for ; Fri, 18 Apr 1997 03:49:57 -0700 (PDT) Received: from donau.informatik.uni-rostock.de (donau [139.30.7.135]) by ceylon.informatik.uni-rostock.de (8.8.5/8.8.5/relay2.1) with ESMTP id MAA16221 for ; Fri, 18 Apr 1997 12:49:47 +0200 Received: (from gunther@localhost) by donau.informatik.uni-rostock.de (8.8.5/8.8.5/fin1.16) id MAA04095 for smp@FreeBSD.org; Fri, 18 Apr 1997 12:49:45 +0200 (MET DST) Date: Fri, 18 Apr 1997 12:49:45 +0200 (MET DST) From: Gunther Hipper Message-Id: <199704181049.MAA04095@donau.informatik.uni-rostock.de> To: smp@FreeBSD.org Subject: Since last sup on April 18th an error in make depend Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: eYdse0dtFBNLK6DY2em4dw== Sender: owner-smp@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Hi ! I'm back again. Yesterday, i successfully compiled and ran the SMP-kernel on a diskless GA-586DX Board with 2 P150 CPUs. Today, i installed a completely new machine on a local EIDE drive (to test the EIDE support, i had some problems yesterday to mount the /dev/wd0a and broke my filesystem..) So, i supped the today in the morning the latest SMP-sys-tree with a: src-sys release=smp host=cvsup.freebsd.org hostbase=/home base=/usr prefix=/usr/sup/cvssmp delete old use-rel-suffix Then, I copied my old SMPFCNA.3 to /usr/src/sys/i386/conf and did a /usr/sbin/config SMPFCNA.3. Config is okay, switch to /usr/src/sys/compile/SMPFCNA.3. make depend and now i got the following error: fcna1:{root} /usr/src/sys-SMP.970418/compile/SMPFCNA.2 [228] make depend cc -c -O -Wreturn-type -Wcomment -Wredundant-decls -Wimplicit -Wnested-externs -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -nostdinc -I- -I. -I../.. -I../../../include -DFAILSAFE -DCOMPAT_43 -DMSDOSFS -DNFS -DFFS -DINET -DKERNEL -DMAXUSERS=10 -UKERNEL ../../i386/i386/genassym.c In file included from ../../i386/i386/genassym.c:71: ../../nfs/nfsdiskless.h:68: field `swap_args' has incomplete type ../../nfs/nfsdiskless.h:75: field `root_args' has incomplete type ../../nfs/nfsdiskless.h:87: field `swap_args' has incomplete type ../../nfs/nfsdiskless.h:93: field `root_args' has incomplete type ../../i386/i386/genassym.c: In function `main': ../../i386/i386/genassym.c:119: `UMAXPTDI' undeclared (first use this function) ../../i386/i386/genassym.c:119: (Each undeclared identifier is reported only once ../../i386/i386/genassym.c:119: for each function it appears in.) ../../i386/i386/genassym.c:119: `UMAXPTEOFF' undeclared (first use this function) *** Error code 1 Stop. I looked for the error, but did'nt find anything yet (e.g. uncommented options NFS). Is this my mistake ? Bye Gunther From owner-freebsd-smp Fri Apr 18 07:08:14 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08088 for smp-outgoing; Fri, 18 Apr 1997 07:08:14 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA08077 for freebsd-smp; Fri, 18 Apr 1997 07:08:09 -0700 (PDT) Date: Fri, 18 Apr 1997 07:08:09 -0700 (PDT) From: Peter Wemm Message-Id: <199704181408.HAA08077@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/conf files newvers.sh sys/i386/boot/biosboot asm.h sys/i386/conf files.i386 options.i386 sys/i386/include clock.h cpu.h cpufunc.h ipl.h pcb.h pmap.h segments.h spl.h sys/miscfs/procfs procfs_status.c sys/net rtsock.c sys/ufs/ufs ufs_inode.c sys/ddb db_trap.c sys/i386/i386 autoconf.c db_interface.c db_trace.c exception.s genassym.c i386-gdbstub.c identcpu.c locore.s machdep.c microtime.s pmap.c support.s swtch.s symbols.raw trap.c sys/i386/isa clock.c icu.h icu.s if_ed.c if_ze.c isa.c isa_device.h npx.c pcibus.c random_machdep.c sio.c vector.s sys/i386/linux linux_misc.c sys/kern init_main.c kern_acct.c kern_clock.c kern_exit.c kern_fork.c kern_ktrace.c kern_resource.c kern_shutdown.c kern_sig.c kern_synch.c subr_prf.c tty.c vfs_bio.c sys/pci pci.c pcibus.h pcireg.h sys/scsi scsi_ioctl.c sd.c sys/sys kernel.h proc.h resourcevar.h sys/vm vm_glue.c vm_meter.c Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 07:08:07 Modified: conf files newvers.sh ddb db_trap.c i386/boot/biosboot asm.h i386/conf files.i386 options.i386 i386/i386 autoconf.c db_interface.c db_trace.c exception.s genassym.c i386-gdbstub.c identcpu.c locore.s machdep.c microtime.s pmap.c support.s swtch.s symbols.raw trap.c i386/include clock.h cpu.h cpufunc.h ipl.h pcb.h pmap.h segments.h spl.h i386/isa clock.c icu.h icu.s if_ed.c if_ze.c isa.c isa_device.h npx.c pcibus.c random_machdep.c sio.c vector.s i386/linux linux_misc.c kern init_main.c kern_acct.c kern_clock.c kern_exit.c kern_fork.c kern_ktrace.c kern_resource.c kern_shutdown.c kern_sig.c kern_synch.c subr_prf.c tty.c vfs_bio.c miscfs/procfs procfs_status.c net rtsock.c pci pci.c pcibus.h pcireg.h scsi scsi_ioctl.c sd.c sys kernel.h proc.h resourcevar.h ufs/ufs ufs_inode.c vm vm_glue.c vm_meter.c Log: First pass of mega -current merge... :-] There are known to be problems at this stage, keep using the old kernel until this compiles again. Revision Changes Path 1.5 +14 -6 sys/conf/files 1.5 +2 -2 sys/conf/newvers.sh 1.5 +12 -10 sys/ddb/db_trap.c 1.6 +1 -1 sys/i386/boot/biosboot/asm.h 1.15 +11 -10 sys/i386/conf/files.i386 1.17 +23 -9 sys/i386/conf/options.i386 1.14 +18 -35 sys/i386/i386/autoconf.c 1.7 +16 -1 sys/i386/i386/db_interface.c 1.4 +15 -9 sys/i386/i386/db_trace.c 1.13 +22 -1 sys/i386/i386/exception.s 1.8 +14 -28 sys/i386/i386/genassym.c 1.3 +8 -0 sys/i386/i386/i386-gdbstub.c 1.6 +290 -7 sys/i386/i386/identcpu.c 1.36 +106 -115 sys/i386/i386/locore.s 1.37 +84 -27 sys/i386/i386/machdep.c 1.18 +1 -1 sys/i386/i386/microtime.s 1.36 +39 -57 sys/i386/i386/pmap.c 1.19 +18 -15 sys/i386/i386/support.s 1.33 +50 -8 sys/i386/i386/swtch.s 1.5 +3 -3 sys/i386/i386/symbols.raw 1.17 +33 -16 sys/i386/i386/trap.c 1.5 +1 -1 sys/i386/include/clock.h 1.10 +5 -1 sys/i386/include/cpu.h 1.15 +15 -1 sys/i386/include/cpufunc.h 1.8 +1 -1 sys/i386/include/ipl.h 1.12 +12 -15 sys/i386/include/pcb.h 1.5 +3 -3 sys/i386/include/pmap.h 1.7 +1 -1 sys/i386/include/segments.h 1.10 +1 -1 sys/i386/include/spl.h 1.15 +25 -12 sys/i386/isa/clock.c 1.15 +1 -1 sys/i386/isa/icu.h 1.23 +1 -1 sys/i386/isa/icu.s 1.8 +3 -3 sys/i386/isa/if_ed.c 1.11 +3 -3 sys/i386/isa/if_ze.c 1.16 +11 -8 sys/i386/isa/isa.c 1.9 +1 -1 sys/i386/isa/isa_device.h 1.17 +1 -2 sys/i386/isa/npx.c 1.8 +24 -11 sys/i386/isa/pcibus.c 1.6 +1 -1 sys/i386/isa/random_machdep.c 1.16 +55 -26 sys/i386/isa/sio.c 1.39 +1 -1 sys/i386/isa/vector.s 1.4 +4 -6 sys/i386/linux/linux_misc.c 1.43 +22 -62 sys/kern/init_main.c 1.4 +20 -21 sys/kern/kern_acct.c 1.5 +13 -2 sys/kern/kern_clock.c 1.12 +6 -7 sys/kern/kern_exit.c 1.10 +53 -36 sys/kern/kern_fork.c 1.7 +5 -5 sys/kern/kern_ktrace.c 1.5 +2 -2 sys/kern/kern_resource.c 1.9 +1 -1 sys/kern/kern_shutdown.c 1.9 +6 -6 sys/kern/kern_sig.c 1.14 +24 -3 sys/kern/kern_synch.c 1.11 +1 -1 sys/kern/subr_prf.c 1.5 +9 -10 sys/kern/tty.c 1.15 +12 -14 sys/kern/vfs_bio.c 1.4 +4 -4 sys/miscfs/procfs/procfs_status.c 1.7 +1 -1 sys/net/rtsock.c 1.9 +13 -4 sys/pci/pci.c 1.7 +1 -1 sys/pci/pcibus.h 1.6 +1 -1 sys/pci/pcireg.h 1.6 +2 -1 sys/scsi/scsi_ioctl.c 1.11 +5 -3 sys/scsi/sd.c 1.14 +3 -3 sys/sys/kernel.h 1.18 +6 -4 sys/sys/proc.h 1.4 +1 -1 sys/sys/resourcevar.h 1.5 +22 -70 sys/ufs/ufs/ufs_inode.c 1.14 +25 -20 sys/vm/vm_glue.c 1.4 +2 -2 sys/vm/vm_meter.c From owner-freebsd-smp Fri Apr 18 07:36:22 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA09657 for smp-outgoing; Fri, 18 Apr 1997 07:36:22 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA09650 for ; Fri, 18 Apr 1997 07:36:15 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id WAA28387; Fri, 18 Apr 1997 22:35:48 +0800 (WST) Message-Id: <199704181435.WAA28387@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Gunther Hipper cc: smp@FreeBSD.org Subject: Re: Since last sup on April 18th an error in make depend In-reply-to: Your message of "Fri, 18 Apr 1997 12:49:45 +0200." <199704181049.MAA04095@donau.informatik.uni-rostock.de> Date: Fri, 18 Apr 1997 22:35:48 +0800 From: Peter Wemm Sender: owner-smp@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Gunther Hipper wrote: [..] > So, i supped the today in the morning the latest SMP-sys-tree with a: > src-sys release=smp host=cvsup.freebsd.org hostbase=/home base=/usr prefix=/u sr/sup/cvssmp delete > old use-rel-suffix [..] > In file included from ../../i386/i386/genassym.c:71: > ../../nfs/nfsdiskless.h:68: field `swap_args' has incomplete type > ../../nfs/nfsdiskless.h:75: field `root_args' has incomplete type > ../../nfs/nfsdiskless.h:87: field `swap_args' has incomplete type > ../../nfs/nfsdiskless.h:93: field `root_args' has incomplete type > ../../i386/i386/genassym.c: In function `main': > ../../i386/i386/genassym.c:119: `UMAXPTDI' undeclared (first use this functio n) > ../../i386/i386/genassym.c:119: (Each undeclared identifier is reported only once > ../../i386/i386/genassym.c:119: for each function it appears in.) > ../../i386/i386/genassym.c:119: `UMAXPTEOFF' undeclared (first use this funct ion) > *** Error code 1 "OOPS!!" Oh dear! It looks like I accidently imported the latest -current into the real tree rather than my copy of it!! UH OH!! Looks like I'd better get on with it and fix it again. :-] The problem you saw was because you have got an mixed up version of the kernel. Some parts are abut 2 months older than others due to the brain-dead design of cvs. I have just committed my merged version so far, but I do not think it'll work yet. There are problems to be resolved with the common_tss code in -current (that I committed to -current, based on somebody else's code.) The changes here are 80% of the way towards being able to have pthreads running via the threading package in libc_r actually executing in parallel on seperate cpu's.... What's remaining is locking on the page tables/ pmaps when a single instance of a process address space is running on two cpu's, and a "driver" engine for libc_r's threads package. (basically, remove the select() based scheduler and use rfork threads instead) Cheers, -Peter From owner-freebsd-smp Fri Apr 18 07:57:54 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA10705 for smp-outgoing; Fri, 18 Apr 1997 07:57:54 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA10697 for freebsd-smp; Fri, 18 Apr 1997 07:57:52 -0700 (PDT) Date: Fri, 18 Apr 1997 07:57:52 -0700 (PDT) From: Peter Wemm Message-Id: <199704181457.HAA10697@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 exception.s Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 07:57:51 Modified: i386/i386 exception.s Log: Fix problem with new code from -current under SMP.. There is one "runtime" per cpu, and it's 8 bytes per entry. Revision Changes Path 1.14 +6 -0 sys/i386/i386/exception.s From owner-freebsd-smp Fri Apr 18 08:01:10 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA10953 for smp-outgoing; Fri, 18 Apr 1997 08:01:10 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA10944 for freebsd-smp; Fri, 18 Apr 1997 08:01:07 -0700 (PDT) Date: Fri, 18 Apr 1997 08:01:07 -0700 (PDT) From: Peter Wemm Message-Id: <199704181501.IAA10944@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 support.s Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 08:01:06 Modified: i386/i386 support.s Log: resolve some unneded divergences from -current - remove str(), it's not used (opposite of ltr()) - load_cr3 comes back. - some minor differences of use of tempoary registers in some wrapper stubs for calling from C code. Revision Changes Path 1.20 +11 -14 sys/i386/i386/support.s From owner-freebsd-smp Fri Apr 18 09:16:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA14348 for smp-outgoing; Fri, 18 Apr 1997 09:16:44 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA14336 for freebsd-smp; Fri, 18 Apr 1997 09:16:40 -0700 (PDT) Date: Fri, 18 Apr 1997 09:16:40 -0700 (PDT) From: Peter Wemm Message-Id: <199704181616.JAA14336@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 machdep.c mp_machdep.c swtch.s trap.c sys/i386/include segments.h Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 09:16:38 Modified: i386/i386 machdep.c mp_machdep.c swtch.s trap.c i386/include segments.h Log: Second pass at merge.. Do some "in theory this might work" tweaks for the per-cpu i386tss structure. This is pretty ugly, the common_tss would be an excellent candidate for the per-cpu private memory pages. Revision Changes Path 1.38 +44 -2 sys/i386/i386/machdep.c 1.36 +7 -6 sys/i386/i386/mp_machdep.c 1.34 +10 -0 sys/i386/i386/swtch.s 1.18 +11 -0 sys/i386/i386/trap.c 1.8 +3 -3 sys/i386/include/segments.h From owner-freebsd-smp Fri Apr 18 09:48:53 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA16930 for smp-outgoing; Fri, 18 Apr 1997 09:48:53 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA16918 for freebsd-smp; Fri, 18 Apr 1997 09:48:48 -0700 (PDT) Date: Fri, 18 Apr 1997 09:48:48 -0700 (PDT) From: Peter Wemm Message-Id: <199704181648.JAA16918@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 symbols.raw Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 09:48:47 Modified: i386/i386 symbols.raw Log: _curpcb doesn't really exist in these kernels Revision Changes Path 1.6 +0 -1 sys/i386/i386/symbols.raw From owner-freebsd-smp Fri Apr 18 21:00:48 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA24133 for smp-outgoing; Fri, 18 Apr 1997 21:00:48 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA24126 for freebsd-smp; Fri, 18 Apr 1997 21:00:45 -0700 (PDT) Date: Fri, 18 Apr 1997 21:00:45 -0700 (PDT) From: Peter Wemm Message-Id: <199704190400.VAA24126@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 swtch.s Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/18 21:00:44 Modified: i386/i386 swtch.s Log: oops.. don't mix up the pointers to the tss Revision Changes Path 1.35 +2 -2 sys/i386/i386/swtch.s From owner-freebsd-smp Fri Apr 18 21:24:01 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id VAA24819 for smp-outgoing; Fri, 18 Apr 1997 21:24:01 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA24801; Fri, 18 Apr 1997 21:23:29 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id MAA00671; Sat, 19 Apr 1997 12:23:16 +0800 (WST) Message-Id: <199704190423.MAA00671@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Peter Wemm cc: freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/i386/i386 swtch.s In-reply-to: Your message of "Fri, 18 Apr 1997 21:00:45 MST." <199704190400.VAA24126@freefall.freebsd.org> Date: Sat, 19 Apr 1997 12:23:16 +0800 From: Peter Wemm Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Peter Wemm wrote: > peter 97/04/18 21:00:44 > > Modified: i386/i386 swtch.s > Log: > oops.. don't mix up the pointers to the tss This stops the fatal page fault on boot, but now it hangs instead.. I'll get more of a chance for a look soon.. Cheers, -Peter From owner-freebsd-smp Sat Apr 19 04:25:02 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id EAA10418 for smp-outgoing; Sat, 19 Apr 1997 04:25:02 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [194.77.0.15]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA10405; Sat, 19 Apr 1997 04:24:44 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id NAA23715; Sat, 19 Apr 1997 13:15:16 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.5/8.8.2) id KAA04308; Sat, 19 Apr 1997 10:03:33 +0200 (CEST) Message-ID: <19970419100332.05559@klemm.gtn.com> Date: Sat, 19 Apr 1997 10:03:32 +0200 From: Andreas Klemm To: Peter Wemm Cc: Peter Wemm , freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/i386/i386 swtch.s References: <199704190400.VAA24126@freefall.freebsd.org> <199704190423.MAA00671@spinner.DIALix.COM> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.69 In-Reply-To: <199704190423.MAA00671@spinner.DIALix.COM>; from Peter Wemm on Sat, Apr 19, 1997 at 12:23:16PM +0800 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 2.2-STABLE Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Apr 19, 1997 at 12:23:16PM +0800, Peter Wemm wrote: > Peter Wemm wrote: > > peter 97/04/18 21:00:44 > > > > Modified: i386/i386 swtch.s > > Log: > > oops.. don't mix up the pointers to the tss > > This stops the fatal page fault on boot, but now it hangs instead.. I'll > get more of a chance for a look soon.. BTW, does this mean, that I can run -current after lite2 merge with all the wonderful aha2940 fixes, if this new kernel version is running ? -- andreas@klemm.gtn.com /\/\___ Wiechers & Partner Datentechnik GmbH Andreas Klemm ___/\/\/ Support Unix -- andreas.klemm@wup.de pgp p-key http://www-swiss.ai.mit.edu/~bal/pks-toplev.html >>> powered by <<< ftp://sunsite.unc.edu/pub/Linux/system/Printing/aps-491.tgz >>> FreeBSD <<< From owner-freebsd-smp Sat Apr 19 04:40:22 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id EAA11465 for smp-outgoing; Sat, 19 Apr 1997 04:40:22 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA11455 for ; Sat, 19 Apr 1997 04:40:11 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id TAA03248; Sat, 19 Apr 1997 19:39:43 +0800 (WST) Message-Id: <199704191139.TAA03248@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Andreas Klemm cc: freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/i386/i386 swtch.s In-reply-to: Your message of "Sat, 19 Apr 1997 10:03:32 +0200." <19970419100332.05559@klemm.gtn.com> Date: Sat, 19 Apr 1997 19:39:42 +0800 From: Peter Wemm Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Andreas Klemm wrote: > On Sat, Apr 19, 1997 at 12:23:16PM +0800, Peter Wemm wrote: > > Peter Wemm wrote: > > > peter 97/04/18 21:00:44 > > > > > > Modified: i386/i386 swtch.s > > > Log: > > > oops.. don't mix up the pointers to the tss > > > > This stops the fatal page fault on boot, but now it hangs instead.. I'll > > get more of a chance for a look soon.. > > BTW, does this mean, that I can run -current after lite2 merge with > all the wonderful aha2940 fixes, if this new kernel version is > running ? Yup. We can probably start thinking about getting the code merged into -current too, once this has settled. Cheers, -Peter From owner-freebsd-smp Sat Apr 19 06:21:17 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA16460 for smp-outgoing; Sat, 19 Apr 1997 06:21:17 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA16453 for freebsd-smp; Sat, 19 Apr 1997 06:21:16 -0700 (PDT) Date: Sat, 19 Apr 1997 06:21:16 -0700 (PDT) From: Peter Wemm Message-Id: <199704191321.GAA16453@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h sys/pc98/i386 locore.s sys/pc98/pc98 ft.c kbd.h sbic55.c sbic55.c.new sbicreg.h sbicvar.h scsireg.h sys/i386/isa if_ix.c if_ixreg.h sys/miscfs/union README libc.fts.c libc.opendir.c libc.readdir.c sys/pci if_pdq.c pdq.c pdq_os.h pdqreg.h sys/vm kern_lock.c lock.h Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/19 06:21:15 Removed: dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h i386/isa if_ix.c if_ixreg.h miscfs/union README libc.fts.c libc.opendir.c libc.readdir.c pc98/i386 locore.s pc98/pc98 ft.c kbd.h sbic55.c sbic55.c.new sbicreg.h sbicvar.h scsireg.h pci if_pdq.c pdq.c pdq_os.h pdqreg.h vm kern_lock.c lock.h Log: Remove defunct files that are no longer in -current. (vm/kern_lock.c is still in -current but shouldn't be, it's unused now that we have kern/kern_lock.c) From owner-freebsd-smp Sat Apr 19 06:20:48 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id GAA16376 for smp-outgoing; Sat, 19 Apr 1997 06:20:48 -0700 (PDT) Received: from corona.jcmax.com (corona.jcmax.com [204.69.248.2]) by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id GAA16367 for ; Sat, 19 Apr 1997 06:20:45 -0700 (PDT) Received: by corona.jcmax.com (5.65/2.49G/4.1.3_U1) id AA18511; Sat, 19 Apr 97 09:20:39 -0400 Date: Sat, 19 Apr 97 09:20:39 -0400 From: cr@jcmax.com (Cyrus Rahman) Message-Id: <9704191320.AA18511@corona.jcmax.com> To: smp@csn.net, smp@freebsd.org Subject: SMP kernel deadlocks Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I've previously described a situation in which the freebsd SMP kernel appeared to deadlock under heavy load. I finally got another chunk of time to look into the problem. **** Problem summary (in Steve's words): Summary of the problem: code: 3-0.970209-SNAP, -current SMP src APIC_IO and all recommended options for same. symptom: heavily loaded system (ie lots of INTs happening) "freezes" reason: cpu0 is trying to service an INT, spin-locks attempting to get the mp_lock, which evidently is permanently held by some process on cpu1. the lock count that is being held is usually 2, but sometimes only 1. reproducing the problem: although I have never seen this before, I can easily reproduce it by disabling the loprio code by changing TEST_LOPRIO to TEST_LOPRIO_NOT in smptests.h. The effect of this is to cause ALL INTs to be serviced by cpu0. **** At the time there was some question about whether there was a true deadlock. As it turns out, there is. The trouble occurs when a page fault occurs on one processor, and, during a critical interval while that page fault is being serviced, an interrupt occurs on the other processor. Defining TEST_LOPRIO decreases the frequency with which this happens, but does not eliminate the problem. The details: During the page fault, it generally happens that at some point smp_invltlb() gets called to flush the TLB on the other CPU's. smp_invltlb() calls allButSelfIPI() and sends an IPI to the other processor, which, unfortunately, is sometimes already processing an interrupt of a higher priority. This interrupt routine now spends its time trying to obtain the mp_lock spin lock so it can enter the kernel, but the processor which has this lock is also in a spin loop in apicIPI() waiting for the IPI to be delivered. Clearly the solution we originally considered, routing the stalled interrupt to the processor with the mp_lock, isn't going to work here. I haven't had time to think through any of the other ways to get around the problem, (and since I need to be in Baltimore in a few hours I probably shouldn't start), but I'd be very interested in any ideas. Cyrus From owner-freebsd-smp Sat Apr 19 07:34:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA19075 for smp-outgoing; Sat, 19 Apr 1997 07:34:44 -0700 (PDT) Received: from nlsystems.com (nlsys.demon.co.uk [158.152.125.33]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA19048; Sat, 19 Apr 1997 07:34:21 -0700 (PDT) Received: from herring.nlsystems.com (herring.nlsystems.com [10.0.0.2]) by nlsystems.com (8.8.5/8.8.5) with SMTP id PAA01083; Sat, 19 Apr 1997 15:34:20 +0100 (BST) Date: Sat, 19 Apr 1997 15:34:20 +0100 (BST) From: Doug Rabson To: Peter Wemm cc: freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h sys/pc98/i386 locore.s sys/pc98/pc98 ft.c kbd.h sbic55.c sbic55.c.new sbicreg.h sbicvar.h scsireg.h sys/i386/isa if_ix.c if_ixreg.h sys/miscfs/union README libc.fts.c libc.opendir.c libc.readdir.c sys/pci if_pdq.c pdq.c pdq_os.h pdqreg.h sys/vm kern_lock.c lock.h In-Reply-To: <199704191321.GAA16453@freefall.freebsd.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 19 Apr 1997, Peter Wemm wrote: > peter 97/04/19 06:21:15 > > Removed: dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h > i386/isa if_ix.c if_ixreg.h > miscfs/union README libc.fts.c libc.opendir.c > libc.readdir.c > pc98/i386 locore.s > pc98/pc98 ft.c kbd.h sbic55.c sbic55.c.new sbicreg.h > sbicvar.h scsireg.h > pci if_pdq.c pdq.c pdq_os.h pdqreg.h > vm kern_lock.c lock.h > Log: > Remove defunct files that are no longer in -current. > (vm/kern_lock.c is still in -current but shouldn't be, it's unused now > that we have kern/kern_lock.c) > Is the SMP kernel supposed to boot yet after the merge? I just tried it and got a panic right after the device probes: lockmgr: not holding exclusive lock lockmgr+0x109 vm_map_lookup+0x41 vm_fault+0x6b trap_pfault+0xe2 ... -- Doug Rabson Mail: dfr@nlsystems.com Nonlinear Systems Ltd. Phone: +44 181 951 1891 From owner-freebsd-smp Sat Apr 19 07:40:14 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA19430 for smp-outgoing; Sat, 19 Apr 1997 07:40:14 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA19421 for ; Sat, 19 Apr 1997 07:40:06 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id WAA05582; Sat, 19 Apr 1997 22:39:44 +0800 (WST) Message-Id: <199704191439.WAA05582@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Doug Rabson cc: freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h sys/pc98/i386 locore.s sys/pc98/pc98 ft.c kbd.h sbic55.c sbic55.c.new sbicreg.h sbicvar.h scsireg.h sys/i386/isa if_ix.c if_ixreg.h sys/miscfs/union README libc.fts.c libc.opendir.c libc.readdir.c sys/pci if_pdq.c pdq.c pdq_os.h pdqreg.h sys/vm kern_lock.c lock.h In-reply-to: Your message of "Sat, 19 Apr 1997 15:34:20 +0100." Date: Sat, 19 Apr 1997 22:39:43 +0800 From: Peter Wemm Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Doug Rabson wrote: > On Sat, 19 Apr 1997, Peter Wemm wrote: > > Remove defunct files that are no longer in -current. > > (vm/kern_lock.c is still in -current but shouldn't be, it's unused now > > that we have kern/kern_lock.c) > > > > Is the SMP kernel supposed to boot yet after the merge? I just tried it > and got a panic right after the device probes: No, but I'm just about out of ideas as to the likely causes. I get a hang when starting scheduling at the moment. I'm dreading the prospect of reviving the serial debug trace code yet again... Cheers, -Peter From owner-freebsd-smp Sat Apr 19 07:42:06 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA19514 for smp-outgoing; Sat, 19 Apr 1997 07:42:06 -0700 (PDT) Received: from ceylon.informatik.uni-rostock.de (ceylon.informatik.uni-rostock.de [139.30.5.237]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA19506 for ; Sat, 19 Apr 1997 07:42:02 -0700 (PDT) Received: from donau.informatik.uni-rostock.de (donau [139.30.7.135]) by ceylon.informatik.uni-rostock.de (8.8.5/8.8.5/relay2.1) with ESMTP id QAA26596; Sat, 19 Apr 1997 16:41:57 +0200 Received: (from gunther@localhost) by donau.informatik.uni-rostock.de (8.8.5/8.8.5/fin1.16) id QAA05093; Sat, 19 Apr 1997 16:41:55 +0200 (MET DST) Date: Sat, 19 Apr 1997 16:41:55 +0200 (MET DST) From: Gunther Hipper Message-Id: <199704191441.QAA05093@donau.informatik.uni-rostock.de> To: smp@csn.net, smp@freebsd.org Subject: SMP-SNAP and EIDE runs fine ! Cc: gunther@informatik.uni-rostock.de Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: GHJU3SkKrWzlaJUavY37iw== Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi all and especially Steve ! And some good news: As to my opinion (or better as to my knowledge), the SMP kernel SNAP (didn't yet try the pacthes from yesterday) now RUNS with an EIDE drive. So, the comment on the EIDE can be removed from the Web page !! EIDE runs fine !!! Bye Gunther Call me an idiot.. but here is a nice story: to overcome the /dev/wd0, I switched to a diskless client. So far. Then, after some work for my boss, I did a complete install of the last SNAP with the non-SMP kernel. Checked the NFS-mounts, my driver, etc. After everything was okay, I decided to switch to the SMP-kernel. Diskless, because of the EIDE problem, and I need this kernel VERY BADLY!! SMP kernel ran fine diskless. Then, I decided to check out the EIDE disk, and did a mount and boot. And here's the story: after creating tars etc. and more and so on, I suddenly remembered that my SWAP SPACE had a local setup on /DEV/WD0S2 !! So, everything is running fine for more than a week ! From owner-freebsd-smp Sat Apr 19 07:48:47 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id HAA19903 for smp-outgoing; Sat, 19 Apr 1997 07:48:47 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA19890 for ; Sat, 19 Apr 1997 07:48:42 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id WAA06082; Sat, 19 Apr 1997 22:48:28 +0800 (WST) Message-Id: <199704191448.WAA06082@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Doug Rabson , freebsd-smp@freefall.freebsd.org Subject: Re: cvs commit: sys/dev/aic7xxx aic7xxx_asm.1 aic7xxx_reg.h ... In-reply-to: Your message of "Sat, 19 Apr 1997 22:39:43 +0800." <199704191439.WAA05582@spinner.DIALix.COM> Date: Sat, 19 Apr 1997 22:48:28 +0800 From: Peter Wemm Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Peter Wemm wrote: > Doug Rabson wrote: > > On Sat, 19 Apr 1997, Peter Wemm wrote: > > > Remove defunct files that are no longer in -current. > > > (vm/kern_lock.c is still in -current but shouldn't be, it's unused now > > > that we have kern/kern_lock.c) > > > > > > > Is the SMP kernel supposed to boot yet after the merge? I just tried it > > and got a panic right after the device probes: > > No, but I'm just about out of ideas as to the likely causes. I get a But on the other hand, a fresh build after removing the junk reports that some files cannot #include now, so I suspect that's a damn good contributing factor for the failures so far.. Cheers, -Peter From owner-freebsd-smp Sat Apr 19 08:14:47 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA21214 for smp-outgoing; Sat, 19 Apr 1997 08:14:47 -0700 (PDT) Received: from ceylon.informatik.uni-rostock.de (ceylon.informatik.uni-rostock.de [139.30.5.237]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA21209 for ; Sat, 19 Apr 1997 08:14:42 -0700 (PDT) Received: from donau.informatik.uni-rostock.de (donau [139.30.7.135]) by ceylon.informatik.uni-rostock.de (8.8.5/8.8.5/relay2.1) with ESMTP id RAA26680; Sat, 19 Apr 1997 17:14:34 +0200 Received: (from gunther@localhost) by donau.informatik.uni-rostock.de (8.8.5/8.8.5/fin1.16) id RAA05104; Sat, 19 Apr 1997 17:14:32 +0200 (MET DST) Date: Sat, 19 Apr 1997 17:14:32 +0200 (MET DST) From: Gunther Hipper Message-Id: <199704191514.RAA05104@donau.informatik.uni-rostock.de> To: smp@csn.net, smp@freebsd.org Subject: Problem with DEC 21050 PCI/PCI bridge Cc: gunther@informatik.uni-rostock.de Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-MD5: 62lS/S6RWdooRaHpjmLeYw== Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi again ! Steve, or someone else with a GA-586DX, please answer: What BIOS-Version are you using on your GA-586DX ? Here is my mptable-output, and it is missing the 21050 PCI-to-PCI-bridge of my network adapter (a four port DEC 21140) ... maybe it is my BIOS ? > It does, and it does support the 3940 IF the motherboard knows how to handle > bridged PCI cards (the 3940 has a PCI bridge chip on it). This motherboard > is know to properly support the 3940 if correctly setup. Check your > BIOS for a setting that describes the MP spec level. It will give you > a choice between version 1.1 and 1.4. Set it to 1.4. Running at > version 1.1 will cause the 3940s to fail miserably. Again a search of the > SMP mail archive for 3940 should provide you with alot of info > on what we did to ensure that they work (and work with shared INTs). At this moment, I'm trying to get my patched driver up and running. It runs just fine when I disable APIC_IO/SMP_INVTBL (although I should'nt disable, I know, and shared IRQs are not the maximum idea). ;-) I also know that there's a problem concerning bridges. Now here is my mptable output and a dmesg of the (running) kernel with APIC_IO/SMP_INVTBL enabled. Result is an /kernel: tansmission timeout (as an interrupt problem). If I disable APIC_IO/SMP_INVTBL, the driver works fine and stable. Could someone give me help or any suggestions for a workaround ? Should I keep on with disabled APIC_IO/SMP_INVTBL or should I (try to) patch the secondary PCI bus into the kernel ? I also tried to set "options NBUS=3" (someone wrote this some time ago). Steve suggested a patch (in a "Re: troubles with smp kernel"). How could I patch the mp_machdep.c for my secondary bus ? Bye Gunther =============================================================================== MPTable, version 2.0.6 ------------------------------------------------------------------------------- MP Floating Pointer Structure: location: BIOS physical address: 0x000f0c80 signature: '_MP_' length: 16 bytes version: 1.1 checksum: 0xf4 mode: Virtual Wire ------------------------------------------------------------------------------- MP Config Table Header: physical address: 0x000f0c94 signature: 'PCMP' base table length: 292 version: 1.1 checksum: 0x31 OEM ID: 'OEM00000' Product ID: 'PROD00000000' OEM table pointer: 0x00000000 OEM table size: 0 entry count: 28 local APIC address: 0xfee00000 extended table length: 0 extended table checksum: 0 ------------------------------------------------------------------------------- MP Config Base Table Entries: -- 1 PCI -- I/O APICs: APIC ID Version State Address 2 0x11 usable 0xfec00000 -- I/O Ints: Type Polarity Trigger Bus ID IRQ APIC ID INT# ExtINT conforms conforms 0 0 2 0 INT conforms conforms 0 1 2 1 INT conforms conforms 0 0 2 2 INT conforms conforms 0 3 2 3 INT conforms conforms 0 4 2 4 INT conforms conforms 0 5 2 5 INT conforms conforms 0 6 2 6 INT conforms conforms 0 7 2 7 INT conforms conforms 0 8 2 8 INT conforms conforms 0 9 2 9 INT conforms conforms 0 10 2 10 INT conforms conforms 0 11 2 11 INT conforms conforms 0 12 2 12 INT conforms conforms 0 13 2 13 INT conforms conforms 0 14 2 14 INT conforms conforms 0 15 2 15 INT active-lo level 1 8:A 2 16 Processors: APIC ID Version State Family Model Step Flags 0 0x11 BSP, usable 5 2 1 0x07bf 1 0x11 AP, usable 5 2 1 0x07bf -- Bus: Bus ID Type 0 ISA 1 PCI -- I/O APICs: APIC ID Version State Address 2 0x11 usable 0xfec00000 -- I/O Ints: Type Polarity Trigger Bus ID IRQ APIC ID INT# ExtINT conforms conforms 0 0 2 0 INT conforms conforms 0 1 2 1 INT conforms conforms 0 0 2 2 INT conforms conforms 0 3 2 3 INT conforms conforms 0 4 2 4 INT conforms conforms 0 5 2 5 INT conforms conforms 0 6 2 6 INT conforms conforms 0 7 2 7 INT conforms conforms 0 8 2 8 INT conforms conforms 0 9 2 9 INT conforms conforms 0 10 2 10 INT conforms conforms 0 11 2 11 INT conforms conforms 0 12 2 12 INT conforms conforms 0 13 2 13 INT conforms conforms 0 14 2 14 INT conforms conforms 0 15 2 15 INT active-lo level 1 8:A 2 16 INT active-lo level 1 9:A 2 17 INT active-lo level 1 10:A 2 18 INT active-lo level 1 12:A 2 19 SMI conforms conforms 0 0 2 23 -- Local Ints: Type Polarity Trigger Bus ID IRQ APIC ID INT# ExtINT active-hi edge 0 0 255 0 NMI active-hi edge 0 0 255 1 ------------------------------------------------------------------------------- # SMP kernel config file options: options SMP # Symmetric MultiProcessor Kernel options APIC_IO # Symmetric (APIC) I/O options NCPU=2 # number of CPUs options NBUS=2 # number of busses options NAPIC=1 # number of IO APICs options NINTR=24 # number of INTs options SMP_INVLTLB # #options SMP_PRIVPAGES # BROKEN, DO NOT use! #options SMP_AUTOSTART # BROKEN, DO NOT use! #options SERIAL_DEBUG # com port debug output =============================================================================== Copyright (c) 1992-1996 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-SMP #0: Sat Apr 19 14:09:18 GMT 1997 root@fcna1.informatik.uni-rostock.de:/usr/src/sys-SMP.geht/compile/SMPFCNA.7 FreeBSD/SMP: Multiprocessor motherboard cpu0 (BSP): apic id: 0, version: 0x00030010 cpu1 (AP): apic id: 1, version: 0x00030010 io0 (APIC): apic id: 2, version: 0x00170011 Calibrating clock(s) relative to mc146818A clock ... i8254 clock: 1193472 Hz CPU: Pentium (586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x3bf real memory = 33554432 (32768K bytes) avail memory = 30322688 (29612K bytes) Probing for devices on PCI bus 0: chip0 rev 1 on pci0:0:0 chip1 rev 1 on pci0:7:0 chip2 rev 0 on pci0:7:1 vga0 rev 227 int a irq 16 on pci0:8:0 Freeing (NOT implimented) irq 11 for ISA cards. chip3 rev 2 on pci0:10:0 ahc0 rev 0 int a irq 19 on pci0:12:0 Freeing (NOT implimented) irq 11 for ISA cards. ahc0: aic7880 Single Channel, SCSI Id=7, 16 SCBs ahc0: waiting for scsi devices to settle scbus0 at ahc0 bus 0 ahc0: Someone reset channel A Probing for devices on PCI bus 1: de0 rev 32 int a irq 11 on pci1:4:0 **** Hi from tulip_read_macaddr !! **** Hi from tulip_read_macaddr TULIP_21140 == 0x3, TULIP_21140A == 0x4, sc->tulip_chipid == 0x4 **** sc->tulip_chipid is fine !! **** This is an Adaptec EM400 Master **** **** EM400 Master returning 0 de0: enabling 100baseTX port de0: Cogent EM400 MS 21140A [10-100Mb/s] pass 2.0 de0: address 00:00:92:95:16:b8 de1 rev 32 int a irq 11 on pci1:5:0 **** Hi from tulip_read_macaddr !! **** Hi from tulip_read_macaddr TULIP_21140 == 0x3, TULIP_21140A == 0x4, sc->tulip_chipid == 0x4 **** sc->tulip_chipid is fine !! **** bcmp(sc->tulip_rombuf+20, testpat_em400ms, 4) failed !! **** This is an EM400 Slave **** root_unit 0 set hwaddr..set boardsw.. de1: enabling 100baseTX port de1: Cogent EM400 SL 21140A [10-100Mb/s] pass 2.0 de1: address 00:00:92:95:16:b9 de2 rev 32 int a irq 11 on pci1:6:0 **** Hi from tulip_read_macaddr !! **** Hi from tulip_read_macaddr TULIP_21140 == 0x3, TULIP_21140A == 0x4, sc->tulip_chipid == 0x4 **** sc->tulip_chipid is fine !! **** bcmp(sc->tulip_rombuf+20, testpat_em400ms, 4) failed !! **** This is an EM400 Slave **** root_unit 1 0 set hwaddr..set boardsw.. de2: enabling 100baseTX port de2: Cogent EM400 SL 21140A [10-100Mb/s] pass 2.0 de2: address 00:00:92:95:16:ba de3 rev 32 int a irq 11 on pci1:7:0 **** Hi from tulip_read_macaddr !! **** Hi from tulip_read_macaddr TULIP_21140 == 0x3, TULIP_21140A == 0x4, sc->tulip_chipid == 0x4 **** sc->tulip_chipid is fine !! **** bcmp(sc->tulip_rombuf+20, testpat_em400ms, 4) failed !! **** This is an EM400 Slave **** root_unit 2 1 0 set hwaddr..set boardsw.. de3: enabling 100baseTX port de3: Cogent EM400 SL 21140A [10-100Mb/s] pass 2.0 de3: address 00:00:92:95:16:bb Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> ed0 at 0x300-0x31f irq 5 on isa ed0: address 00:40:05:32:cb:aa, type NE2000 (16 bit) sio0 at 0x3f8-0x3ff irq 4 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fdc0: NEC 72065B fd0: 1.44MB 3.5in wdc0 at 0x1f0-0x1f7 irq 14 on isa wdc0: unit 0 (wd0): wd0: 2015MB (4127760 sectors), 4095 cyls, 16 heads, 63 S/T, 512 B/S npx0 on motherboard npx0: INT 16 interface Enabled INTs: 1, 2, 3, 4, 5, 6, 7, 8, 11, 14, 19, imen: 0x00f7b601 NFS ROOT: 139.30.4.30:/usr/cl/lo/1 SMP: All idle procs online. SMP: Starting 1st AP! SMP: AP CPU #1 LAUNCHED!! Starting Scheduling... SMP: TADA! CPU #1 made it into the scheduler!. SMP: All 2 CPU's are online! fcna1:{root} /root [50] From owner-freebsd-smp Sat Apr 19 08:38:51 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA22718 for smp-outgoing; Sat, 19 Apr 1997 08:38:51 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA22711 for freebsd-smp; Sat, 19 Apr 1997 08:38:48 -0700 (PDT) Date: Sat, 19 Apr 1997 08:38:48 -0700 (PDT) From: Peter Wemm Message-Id: <199704191538.IAA22711@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/sys kernel.h Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/19 08:38:47 Modified: sys kernel.h Log: hide #include "opt_smp.h" inside #ifdef KERNEL. fixes some usermode compile problems. Revision Changes Path 1.15 +2 -1 sys/sys/kernel.h From owner-freebsd-smp Sat Apr 19 08:41:00 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA22924 for smp-outgoing; Sat, 19 Apr 1997 08:41:00 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA22915 for freebsd-smp; Sat, 19 Apr 1997 08:40:57 -0700 (PDT) Date: Sat, 19 Apr 1997 08:40:57 -0700 (PDT) From: Peter Wemm Message-Id: <199704191540.IAA22915@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/sys proc.h Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/19 08:40:57 Modified: sys proc.h Log: Hide "opt_smp.h" and within #ifdef KERNEL - the includes are not used except in declarations/definitions that are already inside #ifdef KERNEL below, so it makes no sense to make opt_smp.h to be visible to user mode compiles (eg: breaks make world). Revision Changes Path 1.19 +2 -0 sys/sys/proc.h From owner-freebsd-smp Sat Apr 19 08:44:17 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA23090 for smp-outgoing; Sat, 19 Apr 1997 08:44:17 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA23084 for ; Sat, 19 Apr 1997 08:44:08 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id XAA00539 for ; Sat, 19 Apr 1997 23:43:53 +0800 (WST) Message-Id: <199704191543.XAA00539@spinner.DIALix.COM> To: smp@freebsd.org Subject: failed cvs commit message.. Date: Sat, 19 Apr 1997 23:43:52 +0800 From: Peter Wemm Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I botched a remote commit, the original message was: ==== Modified: kern/init_smp.c Log: change #include to - resolves lockmgr() panic! update the idle proc create code to use the "new" style fork that doesn't "return" from the fork() function call - it now boots! (I'm running smp_active=2 right now!) ==== Cheers, -Peter From owner-freebsd-smp Sat Apr 19 08:49:40 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id IAA23225 for smp-outgoing; Sat, 19 Apr 1997 08:49:40 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA23215 for ; Sat, 19 Apr 1997 08:49:36 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id XAA00961; Sat, 19 Apr 1997 23:49:25 +0800 (WST) Message-Id: <199704191549.XAA00961@spinner.DIALix.COM> To: Peter Wemm cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-other@freefall.freebsd.org, smp@freebsd.org Subject: Re: cvs commit: /home/smp/sys/kern init_smp.c In-reply-to: Your message of "Sat, 19 Apr 1997 08:35:16 MST." <199704191535.IAA22509@freefall.freebsd.org> Date: Sat, 19 Apr 1997 23:49:24 +0800 From: Peter Wemm Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Peter Wemm wrote: > peter 97/04/19 08:35:16 > > Modified: home/smp/sys/kern init_smp.c ^^^^^^^^^^^^^^^^ > Log: > change #include to - solves lockmgr panic! > update the idle proc create code to use the "new" style fork that doesn't > "return" from the fork() function call - it now boots! > (I'm running smp_active=2 right now) OOPS! Slip of the fingers with a remote commit.. Sorry, this message went to the wrong commit mailing list.. :-] (Background for those on -committers, the smp kernel was put on hold while the lite2 stuff was being sorted out in -current. It was resync'ed a few days ago, which broke it somewhat.) Cheers, -Peter From owner-freebsd-smp Sat Apr 19 09:04:22 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA24099 for smp-outgoing; Sat, 19 Apr 1997 09:04:22 -0700 (PDT) Received: (from peter@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id JAA24091 for freebsd-smp; Sat, 19 Apr 1997 09:04:19 -0700 (PDT) Date: Sat, 19 Apr 1997 09:04:19 -0700 (PDT) From: Peter Wemm Message-Id: <199704191604.JAA24091@freefall.freebsd.org> To: freebsd-smp Subject: cvs commit: sys/i386/i386 pmap.c Sender: owner-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk peter 97/04/19 09:04:17 Modified: i386/i386 pmap.c Log: cpu_class is declared in Revision Changes Path 1.37 +0 -1 sys/i386/i386/pmap.c From owner-freebsd-smp Sat Apr 19 10:04:53 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA26255 for smp-outgoing; Sat, 19 Apr 1997 10:04:53 -0700 (PDT) Received: from nlsystems.com (nlsys.demon.co.uk [158.152.125.33]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA26239 for ; Sat, 19 Apr 1997 10:04:49 -0700 (PDT) Received: from herring.nlsystems.com (herring.nlsystems.com [10.0.0.2]) by nlsystems.com (8.8.5/8.8.5) with SMTP id SAA00281; Sat, 19 Apr 1997 18:04:21 +0100 (BST) Date: Sat, 19 Apr 1997 18:04:20 +0100 (BST) From: Doug Rabson To: Peter Wemm cc: Peter Wemm , CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org, cvs-other@freefall.freebsd.org, smp@freebsd.org Subject: Re: cvs commit: /home/smp/sys/kern init_smp.c In-Reply-To: <199704191549.XAA00961@spinner.DIALix.COM> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Sat, 19 Apr 1997, Peter Wemm wrote: > Peter Wemm wrote: > > peter 97/04/19 08:35:16 > > > > Modified: home/smp/sys/kern init_smp.c > ^^^^^^^^^^^^^^^^ > > Log: > > change #include to - solves lockmgr panic! > > update the idle proc create code to use the "new" style fork that doesn't > > "return" from the fork() function call - it now boots! > > (I'm running smp_active=2 right now) > > OOPS! Slip of the fingers with a remote commit.. Sorry, this message went > to the wrong commit mailing list.. :-] (Background for those on > -committers, the smp kernel was put on hold while the lite2 stuff was being > sorted out in -current. It was resync'ed a few days ago, which broke it > somewhat.) I must be doing something wrong. I cvsup'ed, updated and built but I still get the lockmgr panic :-(. Are there any pending fixes that I don't have yet maybe? -- Doug Rabson Mail: dfr@nlsystems.com Nonlinear Systems Ltd. Phone: +44 181 951 1891 From owner-freebsd-smp Sat Apr 19 10:28:44 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id KAA27379 for smp-outgoing; Sat, 19 Apr 1997 10:28:44 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA27374 for ; Sat, 19 Apr 1997 10:28:38 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id BAA00592; Sun, 20 Apr 1997 01:28:22 +0800 (WST) Message-Id: <199704191728.BAA00592@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: Doug Rabson cc: smp@freebsd.org Subject: Re: cvs commit: /home/smp/sys/kern init_smp.c In-reply-to: Your message of "Sat, 19 Apr 1997 18:04:20 +0100." Date: Sun, 20 Apr 1997 01:28:21 +0800 From: Peter Wemm Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Doug Rabson wrote: > On Sat, 19 Apr 1997, Peter Wemm wrote: > > > Peter Wemm wrote: > > > peter 97/04/19 08:35:16 > > > > > > Modified: home/smp/sys/kern init_smp.c > > ^^^^^^^^^^^^^^^^ > > > Log: > > > change #include to - solves lockmgr panic! > > > update the idle proc create code to use the "new" style fork that doesn 't > > > "return" from the fork() function call - it now boots! > > > (I'm running smp_active=2 right now) > > > > OOPS! Slip of the fingers with a remote commit.. Sorry, this message went > > to the wrong commit mailing list.. :-] (Background for those on > > -committers, the smp kernel was put on hold while the lite2 stuff was being > > sorted out in -current. It was resync'ed a few days ago, which broke it > > somewhat.) > > I must be doing something wrong. I cvsup'ed, updated and built but I > still get the lockmgr panic :-(. Are there any pending fixes that I don't > have yet maybe? Hmm.. No.. I have 4 other touched files, and they have #ifdef WORK_IN_PROGRESS around the changes, this is not defined. Did you rerun config and "make clean depend"? I ran without APIC_IO and SMP_INVLTLB to start with, and am currently (in X11) running with APIC_IO and SMP_INVLTLB. peter@spinner[1:25am]~/smp/sys-107> uptime 1:25AM up 22 mins, 2 users, load averages: 0.31, 0.26, 0.18 peter@spinner[1:25am]~/smp/sys-108> uname -a FreeBSD spinner.DIALix.COM 3.0-SMP FreeBSD 3.0-SMP #160: Sun Apr 20 00:05:32 WST 1997 peter@spinner.DIALix.COM:/home/peter/smp/sys/compile/SMP i386 The other possibility is the difference in kernel configs.. Oh, are you running P5 or P6? There is a potential problem with tss descriptor caching that might have come out in the P6 under SMP.. > -- > Doug Rabson Mail: dfr@nlsystems.com > Nonlinear Systems Ltd. Phone: +44 181 951 1891 Cheers, -Peter From owner-freebsd-smp Sat Apr 19 11:05:25 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA28798 for smp-outgoing; Sat, 19 Apr 1997 11:05:25 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA28793 for ; Sat, 19 Apr 1997 11:05:21 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id MAA03317; Sat, 19 Apr 1997 12:05:14 -0600 (MDT) Message-Id: <199704191805.MAA03317@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Gunther Hipper cc: smp@freebsd.org Subject: Re: Problem with DEC 21050 PCI/PCI bridge In-reply-to: Your message of "Sat, 19 Apr 1997 17:14:32 +0200." <199704191514.RAA05104@donau.informatik.uni-rostock.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 12:05:14 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I've never had access to a bridge card to check out the GA586DX's behaviour, but am not suprised to find that it fails, almost ALL current generation MP boards do! I just going from memory, but I think this is how I solved the problem for another user. Note that I placed '???' in the code for the actual INT#, as I don't have enough info to decide which it will be (or its just that I haven't had my first latte' yet this morning...) The INT used depends on the slot the card is in. The ahc is not a direct clue as it is a "virtual fifth slot", ie it's builit into the MB and I forget which slot it is shared with. We know it will be one of: INT active-lo level 1 8:A 2 16 INT active-lo level 1 9:A 2 17 INT active-lo level 1 10:A 2 18 INT active-lo level 1 12:A 2 19 ie, INT 16,17,18,19 furthermore we know that whatever slot the vga card is in uses 16: vga0 rev 227 int a irq 16 on pci0:8:0 you could move the vga card around in the empty PCI slots, recording this value each time you boot, and map the slot to APIC INT# values. Or you could just put the de card in the slot currently occuppied by the vga card and use 16 in place of the '???'. Remember that this patch requires that you keep the de card in the same slot once you have the kernel working... my guess is that your vga is in slot 1, that the map will be: slot 1 -> INT16 slot 2 -> INT17 slot 3 -> INT18 slot 4 -> INT19 and that the ahc "shares" slot 4/INT19. let me know how this does/doesn't work... If you map the slots to INTs please forward them here for my rogue database. ------------------------------------ cut -------------------------------------- *** mp_machdep.c.orig Thu Dec 12 01:43:52 1996 --- mp_machdep.c Sat Apr 19 11:44:18 1997 *************** *** 920,932 **** #define SRCBUSDEVICE(I) ((ioApicINTs[(I)].srcBusIRQ >> 2) & 0x1f) #define SRCBUSLINE(I) (ioApicINTs[(I)].srcBusIRQ & 0x03) int ! get_pci_apic_irq( int pciBus __attribute__ ((unused)), ! int pciDevice, int pciInt ) { /** * FIXME: how do we associate a SPECIFIC PCI bus with a unique bus ID??? */ int intr; --pciInt; /* zero based */ --- 920,934 ---- #define SRCBUSDEVICE(I) ((ioApicINTs[(I)].srcBusIRQ >> 2) & 0x1f) #define SRCBUSLINE(I) (ioApicINTs[(I)].srcBusIRQ & 0x03) int ! get_pci_apic_irq( int pciBus, int pciDevice, int pciInt ) { /** * FIXME: how do we associate a SPECIFIC PCI bus with a unique bus ID??? */ int intr; + + if ( (pciBus == 1) && ((pciDevice >= 4) && (pciDevice <=7)) + return ???; --pciInt; /* zero based */ ------------------------------------ cut -------------------------------------- -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 11:19:54 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA29643 for smp-outgoing; Sat, 19 Apr 1997 11:19:54 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA29636 for ; Sat, 19 Apr 1997 11:19:51 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id MAA03367; Sat, 19 Apr 1997 12:19:31 -0600 (MDT) Message-Id: <199704191819.MAA03367@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Peter Wemm cc: Doug Rabson , smp@freebsd.org Subject: Re: cvs commit: /home/smp/sys/kern init_smp.c In-reply-to: Your message of "Sun, 20 Apr 1997 01:28:21 +0800." <199704191728.BAA00592@spinner.DIALix.COM> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 12:19:31 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Peter, > The other possibility is the difference in kernel configs.. Oh, are you > running P5 or P6? There is a potential problem with tss descriptor caching > that might have come out in the P6 under SMP.. I have a P6 system now that I could test this with. I currently am running the 3.0-970209-SNAP world + SMP-current as of march (ie previous to this round of commits) What would be the appropriate steps: cvsup current from freefall make world with UP kernel make UP kernel, install, reboot cvsup SMP make & install new SMP kernel reboot ??? -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 11:24:52 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id LAA29864 for smp-outgoing; Sat, 19 Apr 1997 11:24:52 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA29857 for ; Sat, 19 Apr 1997 11:24:49 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id MAA03391; Sat, 19 Apr 1997 12:24:44 -0600 (MDT) Message-Id: <199704191824.MAA03391@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: cr@jcmax.com (Cyrus Rahman) cc: smp@freebsd.org Subject: Re: SMP kernel deadlocks In-reply-to: Your message of "Sat, 19 Apr 1997 09:20:39 EDT." <9704191320.AA18511@corona.jcmax.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 12:24:44 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Cyrus, good job! I need to think this thru but it sounds plausable. what I don't understand yet is why the cpu is spinning waiting for the the IPI to be delivered. it should just queue up in the local APIC of the target CPU. the only thing I can think of is that the taget APIC's buffer (its fifo is only 2 entries deep) must be full, which would indeed cause what you are describing. I desperately need to finish a job for one of my clients, but will hopefully be able to get to this latter this week. Now that the probelm is identified I don't think the fix will be that hard. thanx again! -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 13:06:36 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id NAA06447 for smp-outgoing; Sat, 19 Apr 1997 13:06:36 -0700 (PDT) Received: from spinner.DIALix.COM (root@spinner.dialix.com [192.203.228.67]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA06442 for ; Sat, 19 Apr 1997 13:06:23 -0700 (PDT) Received: from spinner.DIALix.COM (peter@localhost.DIALix.oz.au [127.0.0.1]) by spinner.DIALix.COM (8.8.5/8.8.5) with ESMTP id EAA01830; Sun, 20 Apr 1997 04:05:51 +0800 (WST) Message-Id: <199704192005.EAA01830@spinner.DIALix.COM> X-Mailer: exmh version 2.0gamma 1/27/96 To: cr@jcmax.com (Cyrus Rahman) cc: smp@csn.net, smp@freebsd.org Subject: Re: SMP kernel deadlocks In-reply-to: Your message of "Sat, 19 Apr 1997 09:20:39 -0400." <9704191320.AA18511@corona.jcmax.com> Date: Sun, 20 Apr 1997 04:05:51 +0800 From: Peter Wemm Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Cyrus Rahman wrote: [...steves summary...] > reason: > cpu0 is trying to service an INT, spin-locks attempting to get the > mp_lock, which evidently is permanently held by some process on cpu1. > the lock count that is being held is usually 2, but sometimes only 1. > The details: > > During the page fault, it generally happens that at some point > smp_invltlb() gets called to flush the TLB on the other CPU's. > smp_invltlb() calls allButSelfIPI() and sends an IPI to the other > processor, which, unfortunately, is sometimes already processing an > interrupt of a higher priority. This interrupt routine now spends > its time trying to obtain the mp_lock spin lock so it can enter the > kernel, but the processor which has this lock is also in a spin loop > in apicIPI() waiting for the IPI to be delivered. Several comments.. First, the IPI's have highest priority from memory.. There shouldn't be any cases where ipi receival is blocked by hardware interrupt priority ordering (I think.. my memory is pretty rough) Second, IPI's are not maskable at the moment.. not even a splhigh().. However, a 'cli' would do it I guess. Third, the FIFO arrangement is being rather poorly used in the present code. From memory again, there is a 2-deep fifo for each hardware priority "level" (level = vector / 16). Since IPI's start at vector#24 (actually ICU_OFFSET + 24, but that doesn't have any effect since ICU_OFFSET is a multiple of 16). This means that the irq 16 -> 23 (generally remapped PCI irq's) are in the same "level" (hence fifo) as the IPI's. I could imagine that it's possible that the PCI interrupts could fill the fifo under heavy load.. That could also explain why I've not seen it here, I have an EISA system that only has irq0->15, so the IPI's have the fifo on that level to themselves. Perhaps we could release the mplock while sending an IPI, and try to grab it back again before continuing... Alternatively, have a timeout on the IPI, and if the apic hasn't recovered after a certain amount of time (ie: it's indefinately "busy"), then release the mplock for a moment and wait and check the status again before refetching the lock. If it still fails to recover, panic rather than hang forever.. Cheers, -Peter From owner-freebsd-smp Sat Apr 19 13:44:45 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id NAA09172 for smp-outgoing; Sat, 19 Apr 1997 13:44:45 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA09167 for ; Sat, 19 Apr 1997 13:44:41 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id OAA03856; Sat, 19 Apr 1997 14:44:24 -0600 (MDT) Message-Id: <199704192044.OAA03856@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Peter Wemm cc: cr@jcmax.com (Cyrus Rahman), smp@freebsd.org Subject: Re: SMP kernel deadlocks In-reply-to: Your message of "Sun, 20 Apr 1997 04:05:51 +0800." <199704192005.EAA01830@spinner.DIALix.COM> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 14:44:24 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Peter, > Several comments.. > ... > Third, the FIFO arrangement is being rather poorly used in the present > code. From memory again, there is a 2-deep fifo for each hardware priority > "level" (level = vector / 16). Since IPI's start at vector#24 (actually > ICU_OFFSET + 24, but that doesn't have any effect since ICU_OFFSET is a > multiple of 16). This means that the irq 16 -> 23 (generally remapped PCI > irq's) are in the same "level" (hence fifo) as the IPI's. > > I could imagine that it's possible that the PCI interrupts could fill the > fifo under heavy load.. That could also explain why I've not seen it > here, I have an EISA system that only has irq0->15, so the IPI's have the > fifo on that level to themselves. this is correct according to what I can remember... --- > Perhaps we could release the mplock while sending an IPI, and try to grab > it back again before continuing... Alternatively, have a timeout on the > IPI, and if the apic hasn't recovered after a certain amount of time (ie: > it's indefinately "busy"), then release the mplock for a moment and wait > and check the status again before refetching the lock. If it still fails > to recover, panic rather than hang forever.. I need to go back and read the intel app notes, etc. to determine the best strategy. I *think* that we could send the IPI, THEN release the mplock, spin till its accepted into a fifo, reclaim the lock and continue. Also remember that the tlbflush IPI is itself incomplete in that it doesn't sync with the other cpu(s), it merely requests the flush then continues without waiting for it to actually occur. When we re-write this IPI correctly the exact method for handling the deadlock should become clearer... The complete re-design of the INTerrupt sub-system can't be far off, we need to redistribute the hardware/software INTs thru-out the entire 256 vector range to properly utilize the APIC structure. And that beast known as vector.s should probably be tossed and redone from the ground up. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 15:18:49 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA18687 for smp-outgoing; Sat, 19 Apr 1997 15:18:49 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA18682 for ; Sat, 19 Apr 1997 15:18:44 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id QAA04153 for ; Sat, 19 Apr 1997 16:18:42 -0600 (MDT) Message-Id: <199704192218.QAA04153@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: smp@freebsd.org Subject: failing PCI bridge cards Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 16:18:42 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I just added a page that attempts to describe the black magic necessary to patch an SMP kernel to use a bridged PCI card: http://www.freebsd.org/~fsmp/SMP/pcibridge.html It still needs work, review/comments appreciated, especially by those whom I have already helped with this particular problem. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 15:34:54 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id PAA19690 for smp-outgoing; Sat, 19 Apr 1997 15:34:54 -0700 (PDT) Received: from cais.cais.com (root@cais.com [199.0.216.4]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA19665 for ; Sat, 19 Apr 1997 15:34:43 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [205.252.122.1]) by cais.cais.com (8.8.5/8.7.3) with SMTP id SAA12335 for ; Sat, 19 Apr 1997 18:34:38 -0400 (EDT) Received: from Journey2.mat.net (journey2.mat.net [205.252.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id SAA01664 for ; Sat, 19 Apr 1997 18:34:37 -0400 Date: Sat, 19 Apr 1997 18:33:52 -0400 (EDT) From: Chuck Robey To: FreeBSD-SMP@FreeBSD.org Subject: ctm of smp-cur Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-smp@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk I see (now that Peter is making changes to the smp code base) that the ctm of the smp cvs tree has halted. I don't have a directly connected site, so I can't do it ... would someone consider restarting the ctm of the smp cvs tree? I have a copy of the old one, I'd sure like the new changes ... ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 9120 Edmonston Ct #302 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-smp Sat Apr 19 18:14:34 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id SAA26595 for smp-outgoing; Sat, 19 Apr 1997 18:14:34 -0700 (PDT) Received: from mail0.iij.ad.jp (mail0.iij.ad.jp [202.232.2.113]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA26588 for ; Sat, 19 Apr 1997 18:14:27 -0700 (PDT) Received: from uucp2.iij.ad.jp (uucp2.iij.ad.jp [202.232.2.202]) by mail0.iij.ad.jp (8.8.5+2.7Wbeta5/3.5Wpl4-MAIL) with SMTP id KAA02945; Sun, 20 Apr 1997 10:13:51 +0900 (JST) Received: (from uucp@localhost) by uucp2.iij.ad.jp (8.6.12+2.4W/3.3W9-UUCP) with UUCP id KAA08703; Sun, 20 Apr 1997 10:13:51 +0900 Received: from tyd1.tydfam.iijnet.or.jp (tyd1.tydfam.iijnet.or.jp [192.168.1.2]) by tydfam.iijnet.or.jp (8.8.5/3.4W2-uucp) with ESMTP id KAA25316; Sun, 20 Apr 1997 10:00:13 +0900 (JST) Received: from localhost.tydfam.iijnet.or.jp (localhost.tydfam.iijnet.or.jp [127.0.0.1]) by tyd1.tydfam.iijnet.or.jp (8.8.5/3.4Wnomx) with SMTP id KAA25887; Sun, 20 Apr 1997 10:00:13 +0900 (JST) Message-Id: <199704200100.KAA25887@tyd1.tydfam.iijnet.or.jp> X-Authentication-Warning: tyd1.tydfam.iijnet.or.jp: localhost.tydfam.iijnet.or.jp [127.0.0.1] didn't use HELO protocol To: peter@spinner.dialix.com Cc: dfr@nlsystems.com, smp@freebsd.org Subject: Q) SMP Kernel compilation Reply-To: ken@tydfam.iijnet.or.jp X-Mailer: Mew version 1.55 on Emacs 19.34.2, Mule 2.3 Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Sun, 20 Apr 1997 10:00:13 +0900 From: Takeshi Yamada Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I have the following compilation error as of Apr. 19th, 2300GMT SMP source. What am I wrong? # make depend; make : : : cc -c -O -Wreturn-type -Wcomment -Wredundant-decls -Wimplicit -Wnested-externs -Wstrict-prototypes -Wmissing-prototypes -Wpointer-arith -nostdinc -I- -I. -I../.. -I../../../include -DSCSI_2_DEF -DEXT2FS -DFDSEEKWAIT=16 -DNSWAPDEV=20 -DMSDOSFS -DMFS -DCD9660 -DNQNFS -DNFS -DFFS -DTCPDEBUG -DIPDIVERT -DTCP_COMPAT_42 -DINET -DMD5 -DCOMPAT_43 -DFAILSAFE -DKERNEL ../../i386/i386/sys_machdep.c ../../i386/i386/sys_machdep.c: In function `i386_get_ldt': ../../i386/i386/sys_machdep.c:158: sizeof applied to an incomplete type ../../i386/i386/sys_machdep.c: In function `i386_set_ldt': ../../i386/i386/sys_machdep.c:225: sizeof applied to an incomplete type *** Error code 1 From owner-freebsd-smp Sat Apr 19 18:29:26 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id SAA27093 for smp-outgoing; Sat, 19 Apr 1997 18:29:26 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA27088 for ; Sat, 19 Apr 1997 18:29:24 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id TAA04732; Sat, 19 Apr 1997 19:29:03 -0600 (MDT) Message-Id: <199704200129.TAA04732@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Peter Wemm cc: Doug Rabson , smp@freebsd.org Subject: Re: cvs commit: /home/smp/sys/kern init_smp.c In-reply-to: Your message of "Sun, 20 Apr 1997 01:28:21 +0800." <199704191728.BAA00592@spinner.DIALix.COM> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 19:29:03 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > Doug Rabson wrote: > > On Sat, 19 Apr 1997, Peter Wemm wrote: > > ... > > I must be doing something wrong. I cvsup'ed, updated and built but I > > still get the lockmgr panic :-(. Are there any pending fixes that I don't > > have yet maybe? > ... > Did you rerun config and "make clean depend"? > > I ran without APIC_IO and SMP_INVLTLB to start with, and am currently (in > X11) running with APIC_IO and SMP_INVLTLB. > ... > The other possibility is the difference in kernel configs.. Oh, are you > running P5 or P6? There is a potential problem with tss descriptor caching > that might have come out in the P6 under SMP.. I supped -current midday today (970419), rebuilt the world, rebuilt the UP kernel, rebooted, worked OK. I then used this -current world to run X11, cvsup cvssmp, build an SMP kernel with APIC_IO and SMP_INVLTLB, rebooted to the SMP kernel, and started the 2nd CPU, worked OK. Am now doing a "make world" with the SMP kernel & both cpus active... will report when it finishes. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Sat Apr 19 18:42:56 1997 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.5/8.8.5) id SAA27515 for smp-outgoing; Sat, 19 Apr 1997 18:42:56 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA27506 for ; Sat, 19 Apr 1997 18:42:52 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.5/8.8.5) with ESMTP id TAA04780; Sat, 19 Apr 1997 19:42:32 -0600 (MDT) Message-Id: <199704200142.TAA04780@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: ken@tydfam.iijnet.or.jp cc: peter@spinner.dialix.com, dfr@nlsystems.com, smp@freebsd.org Subject: Re: Q) SMP Kernel compilation In-reply-to: Your message of "Sun, 20 Apr 1997 10:00:13 +0900." <199704200100.KAA25887@tyd1.tydfam.iijnet.or.jp> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 19 Apr 1997 19:42:32 -0600 Sender: owner-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > I have the following compilation error as of Apr. 19th, 2300GMT > SMP source. What am I wrong? > > # make depend; make have you updated all the rest of your source recently, followed by a "make world"? todays SMP tree is tied to the 3.0-current post Lite-2 merge code. in other words you need to get everything else up to date to be able to use todays SMP code. assumming everything else is current, clean the old objects out first, ie: make clean; make depend; make (in your SMP kernel build directory) -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD