From owner-freebsd-smp Sun Sep 14 00:30:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id AAA28086 for smp-outgoing; Sun, 14 Sep 1997 00:30:17 -0700 (PDT) Received: from sendero-ppp.i-connect.net (sendero-ppp.i-Connect.Net [206.190.143.100]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id AAA28065 for ; Sun, 14 Sep 1997 00:30:11 -0700 (PDT) Received: (qmail 19534 invoked by uid 1000); 14 Sep 1997 07:30:34 -0000 Message-ID: X-Mailer: XFMail 1.2-alpha [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199709140424.VAA07740@usr08.primenet.com> Date: Sun, 14 Sep 1997 00:30:34 -0700 (PDT) Organization: Atlas Telecom From: Simon Shapiro To: Terry Lambert Subject: Re: SMP in FreeBSD 3.x.x Cc: freebsd-smp@freebsd.org, (Paul Missman) Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi Terry Lambert; On 14-Sep-97 you wrote: > > I see that the SMP source was merged with the 3.x.x development > > kernal. > > Does this mean that 3.x.x, when released, will seamlessly support 1 - N > > processors? Also, what do you imagine maximum N will be at 3.x.x > > release > > time? I believe the original Intel spec was up to 4 processors, but I > > see a > > lot of 8 processor servers coming out lately. > > A system can have up to 32 APICs. It is the APIC IDs that limit > the number of processors. > > Given that a system will have an I/O APIC, the max is 31 (except > for I2O systems, where the i960 also has an APIC, or systems with > multiple bus controllers -- each with an APIC). > > So ~30. When I worked at that awful place, we actually computed how many processors one could put on the P6 bus. The number was much, much lower than 30. Our number was based on trying to have the CPUs access memory and I/O. --- Sincerely Yours, (Sent on 14-Sep-97, 00:27:05 by XF-Mail) Simon Shapiro Atlas Telecom Senior Architect 14355 SW Allen Blvd., Suite 130 Beaverton OR 97005 Shimon@i-Connect.Net Voice: 503.643.5559, Emergency: 503.799.2313 From owner-freebsd-smp Sun Sep 14 02:55:53 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id CAA03338 for smp-outgoing; Sun, 14 Sep 1997 02:55:53 -0700 (PDT) Received: from usr06.primenet.com (tlambert@usr06.primenet.com [206.165.6.206]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id CAA03333 for ; Sun, 14 Sep 1997 02:55:51 -0700 (PDT) Received: (from tlambert@localhost) by usr06.primenet.com (8.8.5/8.8.5) id CAA14249; Sun, 14 Sep 1997 02:55:48 -0700 (MST) From: Terry Lambert Message-Id: <199709140955.CAA14249@usr06.primenet.com> Subject: Re: SMP in FreeBSD 3.x.x To: Shimon@i-connect.net (Simon Shapiro) Date: Sun, 14 Sep 1997 09:55:48 +0000 (GMT) Cc: tlambert@primenet.com, freebsd-smp@freebsd.org, missmanp@milo.cfw.com In-Reply-To: from "Simon Shapiro" at Sep 14, 97 00:30:34 am X-Mailer: ELM [version 2.4 PL23] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > When I worked at that awful place, we actually computed how many processors > one could put on the P6 bus. The number was much, much lower than 30. > Our number was based on trying to have the CPUs access memory and I/O. > With respect, that's because the people designing the access paradigms weren't very clever. It's possible to allocate from a global pool to a per processor pool. If I'm trying to allocate out of a per processor pool, then I don't have to contend with other processors to do the allocation. In effect, this is a virtual NUMA. See the Dynix paper: Efficient Kernel Memory Allocation on Shared Memory Multiprocessors McKenney, P.E. and Swignline , J. Proceeding of Winter 1993 Usenix Technical Conference Jan 1993, pages 295-303 Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-smp Sun Sep 14 13:03:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA27287 for smp-outgoing; Sun, 14 Sep 1997 13:03:43 -0700 (PDT) Received: from sendero-ppp.i-connect.net (sendero-ppp.i-Connect.Net [206.190.143.100]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id NAA27272 for ; Sun, 14 Sep 1997 13:03:38 -0700 (PDT) Received: (qmail 23500 invoked by uid 1000); 14 Sep 1997 20:03:59 -0000 Message-ID: X-Mailer: XFMail 1.2-alpha [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199709140955.CAA14249@usr06.primenet.com> Date: Sun, 14 Sep 1997 13:03:59 -0700 (PDT) Organization: Atlas Telecom From: Simon Shapiro To: Terry Lambert Subject: Re: SMP in FreeBSD 3.x.x Cc: missmanp@milo.cfw.com, freebsd-smp@freebsd.org Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi Terry Lambert; On 14-Sep-97 you wrote: > > When I worked at that awful place, we actually computed how many > > processors > > one could put on the P6 bus. The number was much, much lower than 30. > > Our number was based on trying to have the CPUs access memory and I/O. > > > With respect, that's because the people designing the access paradigms > weren't very clever. It's possible to allocate from a global pool > to a per processor pool. > > If I'm trying to allocate out of a per processor pool, then I > don't have to contend with other processors to do the allocation. > > In effect, this is a virtual NUMA. > > See the Dynix paper: > > Efficient Kernel Memory Allocation on Shared Memory Multiprocessors > McKenney, P.E. and Swignline , J. > Proceeding of Winter 1993 Usenix Technical Conference > Jan 1993, pages 295-303 The Dynix papers were used as a model to scale SMP vs. MPP. The concusion (in another job, the awful place did not know their head from their tail :-) was that aroud 30 processors is where scalability will fall off. We used Dynix + Oracle for O/S application model. Processor was Pentium. If I remember correctly, the P6-200 has worse instructions/memory/IO bandwidth ratios than Pentiums-66 does. That led to the conclusion that we will not grow beyond 30 either. I will not go into what the initial P7 was supposed to do. I have no clue where it ended being. --- Sincerely Yours, (Sent on 14-Sep-97, 12:58:09 by XF-Mail) Simon Shapiro Atlas Telecom Senior Architect 14355 SW Allen Blvd., Suite 130 Beaverton OR 97005 Shimon@i-Connect.Net Voice: 503.643.5559, Emergency: 503.799.2313 From owner-freebsd-smp Sun Sep 14 20:53:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id UAA24853 for smp-outgoing; Sun, 14 Sep 1997 20:53:14 -0700 (PDT) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id UAA24848 for ; Sun, 14 Sep 1997 20:53:09 -0700 (PDT) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id UAA24481; Sun, 14 Sep 1997 20:55:03 -0700 (PDT) Message-Id: <199709150355.UAA24481@implode.root.com> To: Simon Shapiro cc: Terry Lambert , missmanp@milo.cfw.com, freebsd-smp@freebsd.org Subject: Re: SMP in FreeBSD 3.x.x In-reply-to: Your message of "Sun, 14 Sep 1997 13:03:59 PDT." From: David Greenman Reply-To: dg@root.com Date: Sun, 14 Sep 1997 20:55:03 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >:-) was that aroud 30 processors is where scalability will fall off. We >used Dynix + Oracle for O/S application model. Processor was Pentium. If >I remember correctly, the P6-200 has worse instructions/memory/IO bandwidth >ratios than Pentiums-66 does. That led to the conclusion that we will not >grow beyond 30 either. I will not go into what the initial P7 was supposed This contradicts a paper that was given at a recent Usenix-sponsored conference that shows that the amount of main memory traffic in P6 systems is *dramatically* reduced over Pentium systems. I forget the exact ratio, but it is something like 1/5th. This is entirely due to the superior cache architecture of the P6. -DG David Greenman Core-team/Principal Architect, The FreeBSD Project From owner-freebsd-smp Sun Sep 14 21:28:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id VAA27199 for smp-outgoing; Sun, 14 Sep 1997 21:28:36 -0700 (PDT) Received: from sendero-ppp.i-connect.net (sendero-ppp.i-Connect.Net [206.190.143.100]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id VAA27187 for ; Sun, 14 Sep 1997 21:28:32 -0700 (PDT) Received: (qmail 29752 invoked by uid 1000); 15 Sep 1997 04:28:53 -0000 Message-ID: X-Mailer: XFMail 1.2-alpha [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199709150355.UAA24481@implode.root.com> Date: Sun, 14 Sep 1997 21:28:53 -0700 (PDT) Organization: Atlas Telecom From: Simon Shapiro To: dg@root.com Subject: Re: SMP in FreeBSD 3.x.x Cc: Terry Lambert , missmanp@milo.cfw.com, freebsd-smp@freebsd.org Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi David Greenman; On 15-Sep-97 you wrote: > >:-) was that aroud 30 processors is where scalability will fall off. We > >used Dynix + Oracle for O/S application model. Processor was Pentium. > >If > >I remember correctly, the P6-200 has worse instructions/memory/IO > >bandwidth > >ratios than Pentiums-66 does. That led to the conclusion that we will > >not > >grow beyond 30 either. I will not go into what the initial P7 was > >supposed > > This contradicts a paper that was given at a recent Usenix-sponsored > conference that shows that the amount of main memory traffic in P6 > systems > is *dramatically* reduced over Pentium systems. I forget the exact > ratio, > but it is something like 1/5th. This is entirely due to the superior > cache > architecture of the P6. For in-cache applications, this is very true. For RDBMS applications this is totally immaterial; In a typical database operation, very little of the data resides in the cache. I hope I am not revealing some proprietary data, but something like an RDBMS requires abot 2MB cache to handle the text. The dataset is so much larger that one can assume the cache does not exist. For example, in a typical RDBMS scenario, one may do a full table scan on several terabytes of data. The cache size (256-512KB) of a P6 is not exactly meaningful here. Also, if you look at the text size of something like Oracle, you will see 8-10MB, The cache represents 2-4% of that (in a P6). Before you mention ordered linking, shared libraries, etc. I'll haste to say that when I last looked, Oracle was statically linked (for a mix of good and bad reasons) and we used to run bets on the number of never-called functions in the code (the real number was about 1,100 or so). Also, the typical RDBMS source is not exactly compiant with KNF and not highly optimized for efficiency. After all, not many FreeBSD'ers work there :-) Although at least one Linux contributor does. I'd better stop now... --- Sincerely Yours, (Sent on 14-Sep-97, 21:19:34 by XF-Mail) Simon Shapiro Atlas Telecom Senior Architect 14355 SW Allen Blvd., Suite 130 Beaverton OR 97005 Shimon@i-Connect.Net Voice: 503.643.5559, Emergency: 503.799.2313 From owner-freebsd-smp Sun Sep 14 21:48:07 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id VAA28224 for smp-outgoing; Sun, 14 Sep 1997 21:48:07 -0700 (PDT) Received: from MindBender.serv.net (mindbender.serv.net [205.153.153.98]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id VAA28209 for ; Sun, 14 Sep 1997 21:48:01 -0700 (PDT) Received: from localhost.HeadCandy.com (localhost.HeadCandy.com [127.0.0.1]) by MindBender.serv.net (8.8.6/8.7.3) with SMTP id VAA04590; Sun, 14 Sep 1997 21:47:35 -0700 (PDT) Message-Id: <199709150447.VAA04590@MindBender.serv.net> X-Authentication-Warning: MindBender.serv.net: localhost.HeadCandy.com [127.0.0.1] didn't use HELO protocol To: Simon Shapiro cc: dg@root.com, Terry Lambert , missmanp@milo.cfw.com, freebsd-smp@freebsd.org Subject: Re: SMP in FreeBSD 3.x.x In-reply-to: Your message of Sun, 14 Sep 97 21:28:53 -0700. Date: Sun, 14 Sep 1997 21:47:30 -0700 From: "Michael L. VanLoon -- HeadCandy.com" Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >Hi David Greenman; On 15-Sep-97 you wrote: >> >:-) was that aroud 30 processors is where scalability will fall off. We >> >used Dynix + Oracle for O/S application model. Processor was Pentium. >> >If >> >I remember correctly, the P6-200 has worse instructions/memory/IO >> >bandwidth >> >ratios than Pentiums-66 does. That led to the conclusion that we will >> >not >> >grow beyond 30 either. I will not go into what the initial P7 was >> >supposed >> This contradicts a paper that was given at a recent Usenix-sponsored >> conference that shows that the amount of main memory traffic in P6 >> systems >> is *dramatically* reduced over Pentium systems. I forget the exact >> ratio, >> but it is something like 1/5th. This is entirely due to the superior >> cache >> architecture of the P6. >For in-cache applications, this is very true. For RDBMS applications this >is totally immaterial; In a typical database operation, very little of the >data resides in the cache. [...] >RDBMS requires abot 2MB cache to handle the text. The dataset is so much [...] >Also, if you look at the text size of something like Oracle, you will see >8-10MB, The cache represents 2-4% of that (in a P6). Before you mention >ordered linking, shared libraries, etc. I'll haste to say that when I last >looked, Oracle was statically linked (for a mix of good and bad reasons) >and we used to run bets on the number of never-called functions in the code >(the real number was about 1,100 or so). [...] Also, remember that not all the world is free Unix (I know you weren't asserting that). There are several commercial software companies that do extensive profiling and reordering of "basic blocks" of compiled code to significantly reduce memory footprint of the binary (and in the course of doing that, improving cache locality). Just because a binary is 8-10MB doesn't mean that it is necessarily accessed "randomly" on all platforms that use the Intel chips. On the other hand, I don't think anyone would deny that a bigger cache is always better, especially with lots of processors. For what it's worth... ----------------------------------------------------------------------------- Michael L. VanLoon michaelv@MindBender.serv.net Contract software development for Windows NT, Windows 95 and Unix. Windows NT and Unix server development in C++ and C. --< Free your mind and your machine -- NetBSD free un*x >-- NetBSD working ports: 386+PC, Mac 68k, Amiga, Atari 68k, HP300, Sun3, Sun4/4c/4m, DEC MIPS, DEC Alpha, PC532, VAX, MVME68k, arm32... NetBSD ports in progress: PICA, others... ----------------------------------------------------------------------------- From owner-freebsd-smp Tue Sep 16 18:00:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id SAA01853 for smp-outgoing; Tue, 16 Sep 1997 18:00:22 -0700 (PDT) Received: from flag.blackened.net (daver@flag.blackened.net [208.206.78.232]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id SAA01835 for ; Tue, 16 Sep 1997 18:00:16 -0700 (PDT) Received: (from daver@localhost) by flag.blackened.net (8.8.7/8.8.5) id SAA29538 for freebsd-smp@freebsd.org; Tue, 16 Sep 1997 18:00:24 -0700 (PDT) Date: Tue, 16 Sep 1997 18:00:24 -0700 (PDT) From: Pomegranate Message-Id: <199709170100.SAA29538@flag.blackened.net> To: freebsd-smp@freebsd.org Subject: Fast Clock with SMP Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk hello. recently set up a SMP box. it has a Tyan Natoma s1686 motherboard, 64mb ram and two PII/266 (intel) cpu's. i slightly modified the SMP-GENERIC kernel, compiled, rebooted, cvsup'd to 3.0-CURRENT and made world. still, since the very first boot, the time-of-day clock has been gaining about 6-7 seconds for every real-time second. i searched the mailing list archives and found one person with the same problem, which apparently occurred while running an SMP kernel in a box with only one CPU. this one has both CPU's, both are seen at bootup by DMESG. --------------------------- FreeBSD 3.0-CURRENT #0: Tue Sep 16 18:22:59 PDT 1997 root@biko.vicor-nb.com:/usr/src/sys/compile/BIKO CPU: Pentium Pro (686-class CPU) Origin = "GenuineIntel" Id = 0x633 Stepping=3 Features=0x80fbff real memory = 67108864 (65536K bytes) avail memory = 62693376 (61224K bytes) FreeBSD/SMP: Multiprocessor motherboard cpu0 (BSP): apic id: 1, version: 0x00040011, at 0xfee00000 cpu1 (AP): apic id: 0, version: 0x00040011, at 0xfee00000 io0 (APIC): apic id: 2, version: 0x00170011, at 0xfec00000 Probing for devices on PCI bus 0: chip0: rev 0x02 on pci0.0.0 chip1: rev 0x01 on pci0.7.0 ide_pci0: rev 0x00 on pci0.7.1 fxp0: rev 0x02 int a irq 18 on pci0.12.0 fxp0: Ethernet address 00:a0:c9:69:02:a3 vga0: rev 0x01 int a irq 17 on pci0.13.0 Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> ed0 not found at 0x280 ed1 not found at 0x300 fe0 not found at 0x300 sio0 at 0x3f8-0x3ff irq 4 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A sio2: disabled, not probed. sio3: disabled, not probed. lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface mse0 at 0x23c irq 5 on isa psm0: disabled, not probed. fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fd0: 1.44MB 3.5in wdc0 at 0x1f0-0x1f7 irq 14 on isa wdc0: unit 0 (wd0): wd0: 3077MB (6303024 sectors), 6253 cyls, 16 heads, 63 S/T, 512 B/S ie0: unknown board_id: f000 ie0 not found at 0x300 ie1: unknown board_id: f000 ie1 not found at 0x360 ep0 not found at 0x300 ex0 not found le0 not found at 0x300 lnc0 not found at 0x280 ze0 not found at 0x300 zp0 not found at 0x300 npx0 on motherboard npx0: INT 16 interface APIC_IO: routing 8254 via 8259 on pin 0 From owner-freebsd-smp Wed Sep 17 08:38:40 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id IAA19228 for smp-outgoing; Wed, 17 Sep 1997 08:38:40 -0700 (PDT) Received: from enst.enst.fr (vt6s9bt1UiMBJcSVzkDhlLNCH4Rj8DJa@enst.enst.fr [137.194.2.16]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id IAA19216 for ; Wed, 17 Sep 1997 08:38:23 -0700 (PDT) Received: from email.enst.fr (email.enst.fr [137.194.168.17]) by enst.enst.fr (8.8.4/8.8.4) with ESMTP id RAA03977 for ; Wed, 17 Sep 1997 17:37:44 +0200 (MET DST) Received: from nikopol.enst.fr (nikopol.enst.fr [137.194.168.105]) by email.enst.fr (8.8.4/8.8.4) with ESMTP id RAA26618; Wed, 17 Sep 1997 17:37:28 +0200 (MET DST) Received: (from fenyo@localhost) by nikopol.enst.fr (8.8.4/8.8.4) id RAA25220; Wed, 17 Sep 1997 17:37:24 +0200 (MET DST) Date: Wed, 17 Sep 1997 17:37:24 +0200 (MET DST) Message-Id: <199709171537.RAA25220@nikopol.enst.fr> From: "Alex Fenyo (eowyn)" To: smp@freebsd.org Subject: ASUS P65UP5 cpu cards question X-WWW: http://home.eowyn.fr.eu.org/~fenyo/documents/axel.html X-PGP-Key: finger alex@eowyn.fr.eu.org X-NIC-Handle: AF713 X-Whois: whois -h whois.internic.net fenyo X-Pager: 06-04-30-75-94 (for emergency only) Organization: Ecole Nationale Superieure des Telecommunications de Paris Reply-to: fenyo@email.enst.fr Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I've seen on the FreeBSD web server that the SMP kernel supports the ASUS P65UP5 dual Pentium board. But I wonder which of the CPU cards have been tested with : - C-P6ND (Pentium Pro) - C-P55T2D (Pentium) Does the FreeBSD SMP kernel support Pentium Pro (maybe it changes nothing ?) ? Many many thanks, Alexandre Fenyö --I can keep a secret, if you can... From owner-freebsd-smp Wed Sep 17 10:12:45 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id KAA25594 for smp-outgoing; Wed, 17 Sep 1997 10:12:45 -0700 (PDT) Received: from pluto.plutotech.com (ken@mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id KAA25586 for ; Wed, 17 Sep 1997 10:12:42 -0700 (PDT) Received: (from ken@localhost) by pluto.plutotech.com (8.8.5/8.8.5) id LAA07897; Wed, 17 Sep 1997 11:12:18 -0600 (MDT) From: Kenneth Merry Message-Id: <199709171712.LAA07897@pluto.plutotech.com> Subject: Re: ASUS P65UP5 cpu cards question In-Reply-To: <199709171537.RAA25220@nikopol.enst.fr> from Alex Fenyo at "Sep 17, 97 05:37:24 pm" To: fenyo@email.enst.fr Date: Wed, 17 Sep 1997 11:12:18 -0600 (MDT) Cc: smp@freebsd.org X-Mailer: ELM [version 2.4ME+ PL28s (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Alex Fenyo wrote... > I've seen on the FreeBSD web server that the SMP kernel supports the > ASUS P65UP5 dual Pentium board. But I wonder which of the CPU cards have > been tested with : > - C-P6ND (Pentium Pro) > - C-P55T2D (Pentium) I'm using the C-P6ND CPU card, and I believe Tor Egge and a few other people are as well. It works just fine. > Does the FreeBSD SMP kernel support Pentium Pro (maybe it changes > nothing ?) ? Yes, the kernel supports the Pentium Pro. Ken -- Kenneth Merry ken@plutotech.com From owner-freebsd-smp Wed Sep 17 12:04:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA02710 for smp-outgoing; Wed, 17 Sep 1997 12:04:14 -0700 (PDT) Received: from sci.brooklyn.cuny.edu (dayton@louis.sci.brooklyn.cuny.edu [146.245.1.7]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA02683 for ; Wed, 17 Sep 1997 12:03:57 -0700 (PDT) Received: (from dayton@localhost) by sci.brooklyn.cuny.edu (8.8.5/8.8.4) id PAA25906; Wed, 17 Sep 1997 15:03:00 -0400 (EDT) Date: Wed, 17 Sep 1997 15:03:00 -0400 (EDT) Message-Id: <199709171903.PAA25906@sci.brooklyn.cuny.edu> From: Dayton Clark To: smp@csn.net CC: smp@freebsd.org In-reply-to: <199708142352.RAA03600@Ilsa.StevesCafe.com> (message from Steve Passe on Thu, 14 Aug 1997 17:52:17 -0600) Subject: Re: 4 CPU machines Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Better late than never. I've been unable to follow the smp list for awhile. Just today I got the 3.0-snap of 7/8/97 working, in SMP mode on an AMI Goliath motherboard with 4 200MHz PPros, two PCI busses and an EISA bus. The disk is an Adaptec 2940. I had to explicitly set NCPU=4 otherwise it would hang during processor discovery. I had thought the default for NCPU was 4, I guess not. I sporadic failure of commands, basic ones make, cmp, ... with an unknown failure, code = 0 message. I was looking through the smp list for any references to this problem when I came upon your request. Is the snap release terribly old? dayton >>>>> "Steve" == Steve Passe writes: Steve> Hi, I would REALLY appreciate hearing from anyone who is Steve> currently running any flavor of 4 CPU SMP kernel. Steve> - P5s or P6s? - motherboard make & model. - disk Steve> controller make & model. - approximate date of last src Steve> update. - anything special that needed to be done to get Steve> it working? Steve> Anyone who attempted it but couldn't get it to work? If Steve> so, with approx. what date src? Steve> -- Steve Passe | powered by smp@csn.net | Symmetric Steve> MultiProcessor FreeBSD From owner-freebsd-smp Wed Sep 17 12:14:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA03358 for smp-outgoing; Wed, 17 Sep 1997 12:14:43 -0700 (PDT) Received: from sci.brooklyn.cuny.edu (dayton@louis.sci.brooklyn.cuny.edu [146.245.1.7]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA03353 for ; Wed, 17 Sep 1997 12:14:36 -0700 (PDT) Received: (from dayton@localhost) by sci.brooklyn.cuny.edu (8.8.5/8.8.4) id PAA25915; Wed, 17 Sep 1997 15:12:35 -0400 (EDT) Date: Wed, 17 Sep 1997 15:12:35 -0400 (EDT) Message-Id: <199709171912.PAA25915@sci.brooklyn.cuny.edu> From: Dayton Clark To: smp@csn.net CC: smp@freebsd.org In-reply-to: <199708141859.MAA02252@Ilsa.StevesCafe.com> (message from Steve Passe on Thu, 14 Aug 1997 12:59:51 -0600) Subject: Re: AMI Goliath motherboard Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I just responded to another message. As of yesterday I have the 3.0-snap of 7/8/97 running with sporadic failure of user commands. dayton >>>>> "Steve" == Steve Passe writes: Steve> Hi, Is anyone using an AMI Goliath motherboard with the SMP Steve> kernel? Has anyone tried to use it but failed? Steve> -- Steve Passe | powered by smp@csn.net | Symmetric Steve> MultiProcessor FreeBSD From owner-freebsd-smp Wed Sep 17 12:58:48 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA05990 for smp-outgoing; Wed, 17 Sep 1997 12:58:48 -0700 (PDT) Received: from ganymede.or.intel.com (root@ganymede.or.intel.com [134.134.248.3]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA05980 for ; Wed, 17 Sep 1997 12:58:42 -0700 (PDT) Received: from ichips.intel.com (ichips.intel.com [134.134.50.200]) by ganymede.or.intel.com (8.8.6/8.8.5) with ESMTP id NAA05959; Wed, 17 Sep 1997 13:01:10 -0700 (PDT) Received: from pdxlx008.jf.intel.com by ichips.intel.com (8.7.4/jIIIa) id MAA15753; Wed, 17 Sep 1997 12:57:51 -0700 (PDT) Received: from pdxlx008 (localhost [127.0.0.1]) by pdxlx008.jf.intel.com (8.8.7/8.7.3) with ESMTP id MAA05986; Wed, 17 Sep 1997 12:56:41 -0700 (PDT) Message-Id: <199709171956.MAA05986@pdxlx008.jf.intel.com> To: dg@root.com cc: Simon Shapiro , Terry Lambert , missmanp@milo.cfw.com, freebsd-smp@freebsd.org Subject: Re: SMP in FreeBSD 3.x.x In-reply-to: Your message of "Sun, 14 Sep 1997 20:55:03 PDT." <199709150355.UAA24481@implode.root.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Date: Wed, 17 Sep 1997 12:56:40 -0700 From: Wayne Scott Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk -----BEGIN PGP SIGNED MESSAGE----- > This contradicts a paper that was given at a recent Usenix-sponsored > conference that shows that the amount of main memory traffic in P6 systems > is *dramatically* reduced over Pentium systems. I forget the exact ratio, > but it is something like 1/5th. This is entirely due to the superior cache > architecture of the P6. For the huge application/poor hit rate case you have to remember that Pentium does not have a write allocate cache. So a write that misses the cache on a Pentium just writes straight to memory. On a Pentium Pro that line is allocated in the L2 possibly evicting another line to make room. This is more traffic if you will never use the data again. (Rare but it does happen with some applications like TPCC) Dr. Nicely's prime number program (of Pentium FDIV fame) thrashes the Pentium Pro's caches nicely this this regard. Note write allocate is a big win for most programs. Wayne Scott MD6 Architecture wscott@ichips.intel.com Work #: (503) 264-4165 Disclaimer: All views expressed are my own opinions, and not necessarily those of Intel Corporation. -----BEGIN PGP SIGNATURE----- Version: 2.6.2 Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface iQCVAwUBNCA188R3RVVTfbqpAQFTnAP/fbxPxFUEjdxuNQOExaU5qLJmzuDq76K6 k5MuoEg2iQpoBnjLckPqMTR0Vb5XMwwcjjzng0XpeO9WMtmA2G2CU2ZJAF9WCN5F zYw5DHQGkHY7N+qw26R/ydwSGvsj+3Wqoi9u+mb9AsT3+5L7QpeWUmwrzTZ0vwqw tfLjpozpd0E= =H2KM -----END PGP SIGNATURE----- From owner-freebsd-smp Thu Sep 18 10:03:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id KAA19593 for smp-outgoing; Thu, 18 Sep 1997 10:03:08 -0700 (PDT) Received: from bermuda.io.com (jher@bermuda.io.com [199.170.88.7]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id KAA19587 for ; Thu, 18 Sep 1997 10:02:55 -0700 (PDT) Received: (from jher@localhost) by bermuda.io.com (8.8.5/8.8.5) id MAA15070 for freebsd-smp@freebsd.org; Thu, 18 Sep 1997 12:02:52 -0500 (CDT) From: jher Message-Id: <199709181702.MAA15070@bermuda.io.com> Subject: ALR Evolution MB To: freebsd-smp@freebsd.org Date: Thu, 18 Sep 1997 12:02:52 -0500 (CDT) X-Mailer: ELM [version 2.4//1 PL25] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I noticed its not listed in the known working hardware. Has anyone tried it? I've got one at work with 2 512kcache PPro 200's in it running the linux SMP kernel (badly) and wanted to see if it would work under freebsd. From owner-freebsd-smp Fri Sep 19 09:00:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id JAA20567 for smp-outgoing; Fri, 19 Sep 1997 09:00:16 -0700 (PDT) Received: from sci.brooklyn.cuny.edu (dayton@louis.sci.brooklyn.cuny.edu [146.245.1.7]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id JAA20561 for ; Fri, 19 Sep 1997 09:00:13 -0700 (PDT) Received: (from dayton@localhost) by sci.brooklyn.cuny.edu (8.8.5/8.8.4) id MAA26122; Fri, 19 Sep 1997 12:00:05 -0400 (EDT) Date: Fri, 19 Sep 1997 12:00:05 -0400 (EDT) Message-Id: <199709191600.MAA26122@sci.brooklyn.cuny.edu> From: Dayton Clark To: smp@freebsd.org Subject: Stable SMP version Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Folks, What is the best starting point for SMP. The web page says 970815-SNAP. Where is this. There is a snap release for 970807, which I have running UP and SMP on my goliath system. User commands sporadically and mysteriously die in SMP mode. So I'm looking for a post-970807 stable version. Any suggestions? thanks dayton From owner-freebsd-smp Fri Sep 19 14:28:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id OAA10111 for smp-outgoing; Fri, 19 Sep 1997 14:28:29 -0700 (PDT) Received: from hutcs.cs.hut.fi (hutcs.cs.hut.fi [130.233.192.7]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id OAA10103 for ; Fri, 19 Sep 1997 14:28:20 -0700 (PDT) Received: (from apl@localhost) by hutcs.cs.hut.fi (8.8.5/8.7.3) id AAA05236; Sat, 20 Sep 1997 00:27:42 +0300 (EET DST) Message-ID: <19970920002741.37382@hutcs.cs.hut.fi> Date: Sat, 20 Sep 1997 00:27:41 +0300 From: Antti-Pekka Liedes To: freebsd-smp@freebsd.org Subject: SMP related problems/weirdness Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.81 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk First of all, deactivating SMP with sysctl -w machdep.smp_active=0 causes the CPU consumption calculations go haywire, ps and top both tell me nothing happens on my computer: last pid: 395; load averages: 0.72, 0.47, 0.21 00:17:34 75 processes: 1 running, 74 sleeping CPU states: 0.0% user, 0.0% nice, 0.0% system, 0.0% interrupt, 0.0% idle Mem: 7328K Active, 48M Inact, 14M Wired, 1468K Cache, 8341K Buf, 21M Free Swap: 243M Total, 128K Used, 242M Free PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND 3 root 28 0 0K 0K psleep 0:00 0.00% 0.00% vmdaemon 4 root 28 0 0K 0K update 0:00 0.00% 0.00% update 282 apl 18 0 632K 796K pause 0:00 0.00% 0.00% zsh 298 apl 18 0 632K 668K pause 0:00 0.00% 0.00% zsh 368 apl 18 0 656K 596K pause 0:00 0.00% 0.00% zsh 324 apl 18 0 656K 556K pause 0:00 0.00% 0.00% zsh etc. Reactivating SMP doesn't repair the situation. Also, the way the CPU percentages of single processes are calculated under SMP seems quite weird. For example, decoding of mp3 stream with mpg123 consumes usually less than 5%, even less than 1% of CPU on my home box (dual P5/200MMX), whereas the same stream consumes about 18% on a Ppro/180. Putting a shell or two into a while;do;done loop also increases the CPU percentages of mpg123, which at least to me seems bogus. -- Antti-Pekka Liedes * apl@IRC * In two hells there's JMT 6 B 406 * apl@iki.fi * one hell too many 02150 ESPOO * apl@apocalypse.tky.hut.fi * - Lucifer +358 - 9 - 468 3121 * +358 - 40 - 5873 593 * (in God's Army) From owner-freebsd-smp Fri Sep 19 20:45:04 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id UAA29989 for smp-outgoing; Fri, 19 Sep 1997 20:45:04 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id UAA29976 for ; Fri, 19 Sep 1997 20:44:59 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id VAA16120; Fri, 19 Sep 1997 21:44:41 -0600 (MDT) Message-Id: <199709200344.VAA16120@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Dayton Clark cc: smp@freebsd.org Subject: Re: Stable SMP version In-reply-to: Your message of "Fri, 19 Sep 1997 12:00:05 EDT." <199709191600.MAA26122@sci.brooklyn.cuny.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 19 Sep 1997 21:44:40 -0600 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > What is the best starting point for SMP. The web page says > 970815-SNAP. Where is this. There is a snap release for 970807, > which I have running UP and SMP on my goliath system. User commands > sporadically and mysteriously die in SMP mode. So I'm looking for a > post-970807 stable version. sometime after 8-30 and b4 the current crop of build problems current.freebsd.org:/pub/FreeBSD -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD