From owner-freebsd-smp Mon Dec 1 07:24:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id HAA14313 for smp-outgoing; Mon, 1 Dec 1997 07:24:38 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from home.dragondata.com (toasty@home.dragondata.com [204.137.237.2]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id HAA14307 for ; Mon, 1 Dec 1997 07:24:33 -0800 (PST) (envelope-from toasty@home.dragondata.com) Received: (from toasty@localhost) by home.dragondata.com (8.8.5/8.8.5) id JAA08978 for freebsd-smp@freebsd.org; Mon, 1 Dec 1997 09:24:27 -0600 (CST) From: Kevin Day Message-Id: <199712011524.JAA08978@home.dragondata.com> Subject: SMP kernel with neptune chipset To: freebsd-smp@freebsd.org Date: Mon, 1 Dec 1997 09:24:27 -0600 (CST) X-Mailer: ELM [version 2.4ME+ PL31 (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 I'm playing with an ancient dual processor pentium motherboard using the neptune chipset. On a uniprocessor kernel, everything runs great. On a SMP kernel any attempt to read/write from a drive on the secondary IDE bus locks up the system. (can't even drop into the debugger) I really am not worried about getting this fixed, as i just put the machine together to run rc564. :) Is it a known problem? Kevin Day From owner-freebsd-smp Mon Dec 1 12:19:11 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA11004 for smp-outgoing; Mon, 1 Dec 1997 12:19:11 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from xoom.com (host074.xoom.com [207.90.142.74]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA10990 for ; Mon, 1 Dec 1997 12:19:03 -0800 (PST) (envelope-from dpsmith@xoom.com) Received: from mail1.xoom.com (host074.xoom.com [207.90.142.74]) by xoom.com (8.8.8/8.8.8) with SMTP id MAA05242; Mon, 1 Dec 1997 12:18:51 -0800 (PST) Date: Mon, 1 Dec 1997 12:18:51 -0800 (PST) From: Dave Smith To: Steve Passe cc: freebsd-smp@freebsd.org Subject: Re: fxp0 or de0 will not work with FreeBSD 3.0 in SMP mode. (fwd) In-Reply-To: <199711301805.LAA13632@Ilsa.StevesCafe.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I cannot find an IO APIC setting in the BIOS. Here is mptable -dmesg output after startup of the SMP kernel. =============================================================================== MPTable, version 2.0.15 ------------------------------------------------------------------------------- MP Floating Pointer Structure: location: BIOS physical address: 0x000f6a50 signature: '_MP_' length: 16 bytes version: 1.1 checksum: 0xe4 mode: Virtual Wire ------------------------------------------------------------------------------- MP Config Table Header: physical address: 0x000f664a signature: 'PCMP' base table length: 244 version: 1.1 checksum: 0xe4 OEM ID: 'OEM00000' Product ID: 'PROD00000000' OEM table pointer: 0x00000000 OEM table size: 0 entry count: 22 local APIC address: 0xfee00000 extended table length: 0 extended table checksum: 0 ------------------------------------------------------------------------------- MP Config Base Table Entries: -- Processors: APIC ID Version State Family Model Step Flags 1 0x11 BSP, usable 6 3 3 0x80fbff 0 0x11 AP, usable 6 3 3 0x80fbff -- Bus: Bus ID Type 0 PCI 1 ISA -- I/O APICs: APIC ID Version State Address 2 0x11 usable 0xfec00000 -- I/O Ints: Type Polarity Trigger Bus ID IRQ APIC ID PIN# ExtINT conforms conforms 1 0 2 0 INT conforms conforms 1 1 2 1 INT conforms conforms 1 0 2 2 INT conforms conforms 1 3 2 3 INT conforms conforms 1 4 2 4 INT conforms conforms 1 5 2 5 INT conforms conforms 1 6 2 6 INT conforms conforms 1 7 2 7 INT conforms conforms 1 8 2 8 INT conforms conforms 1 14 2 14 INT conforms conforms 1 15 2 15 INT active-lo level 1 11 2 16 INT active-lo level 1 10 2 17 INT active-lo level 1 12 2 18 INT active-lo level 1 9 2 19 -- Local Ints: Type Polarity Trigger Bus ID IRQ APIC ID PIN# ExtINT active-hi edge 1 0 255 0 NMI active-hi edge 1 0 255 1 ------------------------------------------------------------------------------- # SMP kernel config file options: # Required: options SMP # Symmetric MultiProcessor Kernel options APIC_IO # Symmetric (APIC) I/O # Optional (built-in defaults will work in most cases): #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 ------------------------------------------------------------------------------- dmesg output: Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-971125-SNAP #0: Wed Nov 26 04:47:17 PST 1997 root@www6.xoom.com:/usr/src/sys/compile/MYKERNEL-SMP CPU: Pentium Pro (686-class CPU) Origin = "GenuineIntel" Id = 0x633 Stepping=3 Features=0x80fbff real memory = 268435456 (262144K bytes) avail memory = 258609152 (252548K 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 0x03 on pci0.0.0 chip1: rev 0x03 on pci0.1.0 chip2: rev 0x01 on pci0.4.0 ide_pci0: rev 0x01 on pci0.4.1 chip3: rev 0x01 int d irq 9 on pci0.4.2 chip4: rev 0x01 on pci0.4.3 vga0: rev 0x44 int a irq 12 on pci0.10.0 fxp0: rev 0x02 int a irq 10 on pci0.11.0 fxp0: Ethernet address 00:a0:c9:8a:ba:59 fxp1: rev 0x02 int a irq 11 on pci0.12.0 fxp1: Ethernet address 00:a0:c9:8a:ba:3e Probing for devices on PCI bus 1: Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> 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: FIFO enabled, 8 bytes threshold fd0: 1.44MB 3.5in wdc0 at 0x1f0-0x1f7 irq 14 on isa wdc0: unit 0 (wd0): wd0: 1907MB (3907008 sectors), 3876 cyls, 16 heads, 63 S/T, 512 B/S npx0 on motherboard npx0: INT 16 interface APIC_IO: routing 8254 via 8259 on pin 0 SMP: AP CPU #1 Launched! fxp0: device timeout =============================================================================== On Sun, 30 Nov 1997, Steve Passe wrote: > Hi, > > Attached is the output of mptable -dmesg. > > Also attached is my smp kernel config file which is the same on both > > machines. > > (The machine was booted into the generic kernel on 3.0 so I could grab > > these attachments.) > ... > > CPU: Pentium Pro (686-class CPU) > > Origin = "GenuineIntel" Id = 0x633 Stepping=3 > > Features=0x80fbff > > real memory = 268435456 (262144K bytes) > > avail memory = 258609152 (252548K 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 > > this looks like it must be an SMP kernel. > > the output from dmesg suggests that the IO APIC is not enabled in the BIOS. > > -- > Steve Passe | powered by > smp@csn.net | Symmetric MultiProcessor FreeBSD > > > From owner-freebsd-smp Mon Dec 1 12:25:52 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA11781 for smp-outgoing; Mon, 1 Dec 1997 12:25:52 -0800 (PST) (envelope-from owner-freebsd-smp) 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 MAA11764 for ; Mon, 1 Dec 1997 12:25:47 -0800 (PST) (envelope-from smp@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id NAA00395; Mon, 1 Dec 1997 13:25:24 -0700 (MST) Message-Id: <199712012025.NAA00395@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Dave Smith cc: freebsd-smp@freebsd.org Subject: Re: fxp0 or de0 will not work with FreeBSD 3.0 in SMP mode. (fwd) In-reply-to: Your message of "Mon, 01 Dec 1997 12:18:51 PST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 01 Dec 1997 13:25:23 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > I cannot find an IO APIC setting in the BIOS. > Here is mptable -dmesg output after startup of the SMP kernel. > ... unfortunately the mptable is usually static info in the BIOS, and won't look any different whether the IO APIC is enabled or not. The suspicious lines are: >fxp0: rev 0x02 int a irq 10 on pci0.11.0 >fxp1: rev 0x02 int a irq 11 on pci0.12.0 according to the mptable: > INT active-lo level 1 11 2 16 > INT active-lo level 1 10 2 17 these should show up on INTs 17 & 16, NOT 10 & 11, if the IO APIC is enabled. Look for options like MP version 1.4 -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Mon Dec 1 12:48:09 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA15145 for smp-outgoing; Mon, 1 Dec 1997 12:48:09 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from pluto.plutotech.com (mail.plutotech.com [206.168.67.137]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA15124 for ; Mon, 1 Dec 1997 12:48:04 -0800 (PST) (envelope-from ken@plutotech.com) Received: (from ken@localhost) by pluto.plutotech.com (8.8.7/8.8.5) id NAA21894; Mon, 1 Dec 1997 13:47:44 -0700 (MST) From: Kenneth Merry Message-Id: <199712012047.NAA21894@pluto.plutotech.com> Subject: Re: fxp0 or de0 will not work with FreeBSD 3.0 in SMP mode. (fwd) In-Reply-To: from Dave Smith at "Dec 1, 97 12:18:51 pm" To: dpsmith@xoom.com (Dave Smith) Date: Mon, 1 Dec 1997 13:47:44 -0700 (MST) Cc: smp@csn.net, freebsd-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 Dave Smith wrote... > > I cannot find an IO APIC setting in the BIOS. > Here is mptable -dmesg output after startup of the SMP kernel. > > > =============================================================================== > > MPTable, version 2.0.15 > > ------------------------------------------------------------------------------- > > MP Floating Pointer Structure: > > location: BIOS > physical address: 0x000f6a50 > signature: '_MP_' > length: 16 bytes > version: 1.1 > checksum: 0xe4 > mode: Virtual Wire Have you tried changing the MP Spec version in your BIOS from 1.1 to 1.4? I've got an ASUS P/I-P65UP5 w/ C-P6ND, and, IIRC, it won't work right without MP Spec version 1.4 enabled in the BIOS. Ken -- Kenneth Merry ken@plutotech.com From owner-freebsd-smp Mon Dec 1 19:32:54 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id TAA21106 for smp-outgoing; Mon, 1 Dec 1997 19:32:54 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from xoom.com (host074.xoom.com [207.90.142.74]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id TAA21101 for ; Mon, 1 Dec 1997 19:32:52 -0800 (PST) (envelope-from dpsmith@xoom.com) Received: from mail1.xoom.com (host074.xoom.com [207.90.142.74]) by xoom.com (8.8.8/8.8.8) with SMTP id TAA09869; Mon, 1 Dec 1997 19:32:33 -0800 (PST) Date: Mon, 1 Dec 1997 19:32:33 -0800 (PST) From: Dave Smith To: Kenneth Merry cc: smp@csn.net, freebsd-smp@freebsd.org Subject: Re: fxp0 or de0 will not work with FreeBSD 3.0 in SMP mode. (fwd) In-Reply-To: <199712012047.NAA21894@pluto.plutotech.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I will check this out tomorrow. I moved the machines to another location which I will have access to tomorrow. I will enable 'MPS 1.4 Support' if it is disabled and hopefully it will work thanks to all your help on this great list. Thanks, Dave. On Mon, 1 Dec 1997, Kenneth Merry wrote: > Dave Smith wrote... > > > > I cannot find an IO APIC setting in the BIOS. > > Here is mptable -dmesg output after startup of the SMP kernel. > > > > > > =============================================================================== > > > > MPTable, version 2.0.15 > > > > ------------------------------------------------------------------------------- > > > > MP Floating Pointer Structure: > > > > location: BIOS > > physical address: 0x000f6a50 > > signature: '_MP_' > > length: 16 bytes > > version: 1.1 > > checksum: 0xe4 > > mode: Virtual Wire > > > Have you tried changing the MP Spec version in your BIOS from 1.1 > to 1.4? I've got an ASUS P/I-P65UP5 w/ C-P6ND, and, IIRC, it won't work > right without MP Spec version 1.4 enabled in the BIOS. > > > Ken > -- > Kenneth Merry > ken@plutotech.com > From owner-freebsd-smp Wed Dec 3 09:02:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id JAA24888 for smp-outgoing; Wed, 3 Dec 1997 09:02:42 -0800 (PST) (envelope-from owner-freebsd-smp) 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 JAA24882 for ; Wed, 3 Dec 1997 09:02:37 -0800 (PST) (envelope-from fbsd@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id JAA08702 for ; Wed, 3 Dec 1997 09:57:24 -0700 (MST) Message-Id: <199712031657.JAA08702@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: smp@freebsd.org Subject: SMP Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 09:57:24 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I've taken the month of December off, and hope to apply much of it to SMP work. I'd like to attack the following: 1: fix any true bugs that have crept in, particularily those preventing anyone from running a specific hardware setup. 2: finish my 1st stage lock-pushdown. I think I'm only a bug or 2 from getting this to run... 3: begin the design of "the real thing". My current lock-pushdown attempts to co-exist with the splxxx paradigm. I'm pretty much convinced at this point that the work I've done in this area is only useful to prove it ain't gonna' cut it. We need to design IN DETAIL a shift to a mutex based kernel. One obvious question is whether we move both UP and SMP that direction, or just SMP. There are many,many other questions to be answered. It would be nice if we could progress on this issue in a serious manner, getting a design wrapped up b4 I have to go back to my real job... -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Wed Dec 3 15:16:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA21863 for smp-outgoing; Wed, 3 Dec 1997 15:16:29 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mailhost.Ipsilon.COM (mailhost.ipsilon.com [205.226.5.12]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id PAA21847 for ; Wed, 3 Dec 1997 15:16:24 -0800 (PST) (envelope-from jre@ipsilon.com) Received: from radio.ipsilon.com (radio.Ipsilon.COM [205.226.28.3]) by mailhost.Ipsilon.COM (8.6.11/8.6.10) with SMTP id PAA00299; Wed, 3 Dec 1997 15:14:47 -0800 Message-ID: <3485E7F5.15FB7483@ipsilon.com> Date: Wed, 03 Dec 1997 15:15:01 -0800 From: Joe Eykholt Organization: Ipsilon Software Engineering X-Mailer: Mozilla 3.01Gold (X11; I; FreeBSD 2.1.0-RELEASE i386) MIME-Version: 1.0 To: Steve Passe CC: smp@freebsd.org Subject: Re: SMP References: <199712031657.JAA08702@Ilsa.StevesCafe.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Steve, You wrote: > 3: begin the design of "the real thing". My current lock-pushdown attempts > to co-exist with the splxxx paradigm. I'm pretty much convinced at this > point that the work I've done in this area is only useful to prove it > ain't gonna' cut it. We need to design IN DETAIL a shift to a mutex based > kernel. One obvious question is whether we move both UP and SMP that > direction, or just SMP. There are many,many other questions to be > answered. > It would be nice if we could progress on this issue in a serious manner, > getting a design wrapped up b4 I have to go back to my real job... I've been playing with the SMP code a bit. I have some design suggestions (and some code snippets) that might or might not be useful. Some you've probably already got lots of ideas about. For what they're worth: 1. I agree that the mutex_lock/unlock based approach would be nice, and more explicit, and possibly better even in the UP case. Hopefully a persuasive design document can convince people it's worth the pain. It should be possible to have spin-type mutexes directly correspond to the various splxxx routines at first, and then break them up into finer-grained locks. I'd use the pthreads_mutex_lock() interfaces (except perhaps leave the pthreads_ prefix off of the names). The locks would be initialized with information about which interrupts are automatically blocked while the lock is held, etc. BTW, I don't think locks should be allowed to be recursively grabbed. 2. Whether a first-level interrupt handler automatically gets a lock blocking all similarly-registered (same imask) interrupts or not is an area to consider. This is analogous to raising CPL before re-enabling. I'd like to investigate very different interrupt models, where interrupts are scheduled as separate threads, similar to but not exactly like Solaris. I think this is possible without massive driver changes. 3. I'd change the CPU-private variables to be inside a structure (struct cpu). There should be two structures, actually, one which is portable and one which is machine-dependent. The portable one can be inside the machine-dependent one, or vice-versa. This way, per-CPU variables are easier to identify. (e.g. cpl and ipending are per-CPU, and seeing CPU->cpu_cpl makes this more clear. 4. The per-CPU mapping causes problems with rfork and with multi-threading in general, and may also hurt context-switching. I prefer the approach of reserving a segment and a segment register (%gs) to select the per-CPU structure throughout the kernel. Unfortunately this does require loading %gs using the local APIC ID on every kernel entry. It does make some accesses somewhat more expensive ... I'm not sure how much. Solaris does this (which is what made me think of it). On non-i386 architectures, there's usually some other register available which the compiler doesn't use (%g7 on SPARC, %r2 on PowerPC) that can point to the per-CPU or per-thread data. Actually, pointing at the per-thread (or per-process until there are kernel threads) data rather than per-CPU data is better in the long term. This is because then preemption isn't a problem ... you can be preempted and moved to another CPU and your per-thread data is still in the same place. You can find the current CPU pointer through your per-thread data. Inlining references through the %gs register can actually reduce code size from the current curproc method. 5. The APIC vectors need to be re-arranged into a priority order so that interrupts don't need to access the I/O APIC to mask off the interrupt during handling. (Or maybe there's a better way to do this). I noticed that level-sensitive interrupts are getting taken twice, because the first interrupt only masks with CPL, so after the EOI and sti, the interrupt is still pending and is taken again. Only the second time does it get masked in the I/O APIC. Being a central resource, locking the IOAPIC on every interrupt is unacceptable, so restructuring the interrupts in a priority order and deferring the EOI until the end seems necessary. I don't completely like this, so I'm hoping there's a better way. If you'd like to discuss any of these areas further, I'd be happy to. I have some code developed around issues #3 and #4, which you could have. Thanks, Joe Eykholt From owner-freebsd-smp Wed Dec 3 15:48:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA24271 for smp-outgoing; Wed, 3 Dec 1997 15:48:01 -0800 (PST) (envelope-from owner-freebsd-smp) 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 PAA24236 for ; Wed, 3 Dec 1997 15:47:51 -0800 (PST) (envelope-from smp@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id QAA10070; Wed, 3 Dec 1997 16:47:45 -0700 (MST) Message-Id: <199712032347.QAA10070@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Joe Eykholt cc: smp@freebsd.org Subject: Re: SMP In-reply-to: Your message of "Wed, 03 Dec 1997 15:15:01 PST." <3485E7F5.15FB7483@ipsilon.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 16:47:45 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > 5. The APIC vectors need to be re-arranged into a priority order so > that interrupts > don't need to access the I/O APIC to mask off the interrupt during > handling. > (Or maybe there's a better way to do this). I noticed that > level-sensitive > interrupts are getting taken twice, because the first interrupt only > masks with CPL, > so after the EOI and sti, the interrupt is still pending and is taken > again. Only > the second time does it get masked in the I/O APIC. Being a central sounds like you are describing our "lazy masking" an attempt to skip blocking an INT unless we get hit a second time. If what you are saying about the second hit being automatic is true, its something for me to tackle real soon. Have you actually demonstrated this to occur all times? Or just ocassionally? OR just via code examination? Could you descibe a specific example that would help me grok the specific situation? > resource, locking > the IOAPIC on every interrupt is unacceptable, so restructuring the > interrupts in > a priority order and deferring the EOI until the end seems necessary. I > don't > completely like this, so I'm hoping there's a better way. I think INT priority ordering will need to be reserved for "execution priority" as oppossed to critical region protection. I'd like it to be dynamic, ie priorities set at boot time as oppossed to compile time. I definately like the idea of using kernel threads to execute ISRs, as oppossed to the current model of running them in the context of whatever process gets hit by the INT. Just not sure how pratical this would be. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Wed Dec 3 16:25:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA26902 for smp-outgoing; Wed, 3 Dec 1997 16:25:43 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mailhost.Ipsilon.COM (mailhost.ipsilon.com [205.226.5.12]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id QAA26893 for ; Wed, 3 Dec 1997 16:25:38 -0800 (PST) (envelope-from jre@ipsilon.com) Received: from radio.ipsilon.com (radio.Ipsilon.COM [205.226.28.3]) by mailhost.Ipsilon.COM (8.6.11/8.6.10) with SMTP id QAA02870; Wed, 3 Dec 1997 16:24:34 -0800 Message-ID: <3485F851.3F54BC7E@ipsilon.com> Date: Wed, 03 Dec 1997 16:24:49 -0800 From: Joe Eykholt Organization: Ipsilon Software Engineering X-Mailer: Mozilla 3.01Gold (X11; I; FreeBSD 2.1.0-RELEASE i386) MIME-Version: 1.0 To: Steve Passe CC: smp@freebsd.org Subject: Re: SMP References: <199712032347.QAA10070@Ilsa.StevesCafe.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Steve Passe wrote: > > Hi, > > > 5. The APIC vectors need to be re-arranged into a priority order so > > that interrupts > > don't need to access the I/O APIC to mask off the interrupt during > > handling. > > (Or maybe there's a better way to do this). I noticed that > > level-sensitive > > interrupts are getting taken twice, because the first interrupt only > > masks with CPL, > > so after the EOI and sti, the interrupt is still pending and is taken > > again. Only > > the second time does it get masked in the I/O APIC. Being a central > > sounds like you are describing our "lazy masking" an attempt to skip blocking > an INT unless we get hit a second time. If what you are saying about the > second hit being automatic is true, its something for me to tackle real soon. > Have you actually demonstrated this to occur all times? Or just ocassionally? > OR just via code examination? Could you descibe a specific example that would > help me grok the specific situation? I think so. I added a counter for interrupts deferred due to cpl and that counter was 1/2 the total interrupt count (I counted all interrupts whether deferred or not). All my interrupts were PCI devices (level sensitive), so I'm pretty sure that's what's happening. I also saw a stack trace that showed that the second interrupt occurred just after the first handler did the sti instruction. I've made other changes to the INTR() macro, but I think the same thing would happen in the -current APIC version. Joe From owner-freebsd-smp Wed Dec 3 16:28:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA27070 for smp-outgoing; Wed, 3 Dec 1997 16:28:29 -0800 (PST) (envelope-from owner-freebsd-smp) 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 QAA27064 for ; Wed, 3 Dec 1997 16:28:26 -0800 (PST) (envelope-from smp@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id RAA10221; Wed, 3 Dec 1997 17:28:23 -0700 (MST) Message-Id: <199712040028.RAA10221@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Joe Eykholt cc: smp@freebsd.org Subject: Re: SMP In-reply-to: Your message of "Wed, 03 Dec 1997 16:24:49 PST." <3485F851.3F54BC7E@ipsilon.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 17:28:22 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > > Have you actually demonstrated this to occur all times? Or just ocassionally? > > OR just via code examination? Could you descibe a specific example that would > > help me grok the specific situation? > > I think so. I added a counter for interrupts deferred due to cpl and > that > counter was 1/2 the total interrupt count (I counted all interrupts > whether > deferred or not). All my interrupts were PCI devices (level sensitive), > so > I'm pretty sure that's what's happening. I also saw a stack trace that > showed > that the second interrupt occurred just after the first handler did the > sti > instruction. > > I've made other changes to the INTR() macro, but I think the same thing > would > happen in the -current APIC version. thanx, I'll look at this tomorrow... -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Wed Dec 3 17:02:07 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA29739 for smp-outgoing; Wed, 3 Dec 1997 17:02:07 -0800 (PST) (envelope-from owner-freebsd-smp) 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 RAA29724 for ; Wed, 3 Dec 1997 17:02:01 -0800 (PST) (envelope-from smp@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id SAA10349; Wed, 3 Dec 1997 18:01:58 -0700 (MST) Message-Id: <199712040101.SAA10349@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Joe Eykholt cc: smp@freebsd.org Subject: Re: SMP In-reply-to: Your message of "Wed, 03 Dec 1997 16:24:49 PST." <3485F851.3F54BC7E@ipsilon.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 18:01:57 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Joe, I think I see the "double level INT" problem: line 242 of i386/isa/apic_vector.s: movl $0, lapic_eoi ; /* XXX too soon? */ \ incb _intr_nesting_level ; \ ; \ /* entry point used by doreti_unpend for HWIs. */ \ __CONCAT(Xresume,irq_num): ; \ --- as you can see I already recognized this as a problem, but didn't have an easy answer. The lapic_eoi should NOT really be sent until after the hardware specific code has been run. At that point the INT source will have been cleared, and it would be safe to send the lapic_eoi, which tells the IO apic that if the level is still asserted it should send another INT. The problem is that we can't place in in the code at the proper place: line 258: ; \ pushl _intr_unit + (irq_num) * 4 ; \ incl _inside_intr ; \ sti ; \ call *_intr_handler + (irq_num) * 4 ; \ cli ; \ decl _inside_intr ; \ ; proper place for lapic_eoi (ie NOT in current): movl $0, lapic_eoi ; we can't do this since this code is sometimes (often) run from doreti as oppossed to the ISR that was invoked when the INT occured. Its already done when the ISR discovers that it can't proceed, and masks the INT at line 276, followed by the lapic_eoi. Doing it again from doreti would be a problem. I think the trick is to modify UNMASK_IRQ(irq_num) to test for level INTs when unmasking an INT. If its a level INT send the lapic_eoi then. If I send you a patch to current apic_vector.s as of earlier today would you be able to run your test code on it? -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Wed Dec 3 17:44:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA03592 for smp-outgoing; Wed, 3 Dec 1997 17:44:26 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mailhost.Ipsilon.COM (mailhost.ipsilon.com [205.226.5.12]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id RAA03587 for ; Wed, 3 Dec 1997 17:44:23 -0800 (PST) (envelope-from jre@ipsilon.com) Received: from radio.ipsilon.com (radio.Ipsilon.COM [205.226.28.3]) by mailhost.Ipsilon.COM (8.6.11/8.6.10) with SMTP id RAA05088; Wed, 3 Dec 1997 17:27:33 -0800 Message-ID: <34860715.31DFF4F5@ipsilon.com> Date: Wed, 03 Dec 1997 17:27:49 -0800 From: Joe Eykholt Organization: Ipsilon Software Engineering X-Mailer: Mozilla 3.01Gold (X11; I; FreeBSD 2.1.0-RELEASE i386) MIME-Version: 1.0 To: Steve Passe CC: smp@freebsd.org Subject: Re: SMP References: <199712040101.SAA10349@Ilsa.StevesCafe.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Steve, I'm glad you see the problem. I wasn't pressing for an immediate solution so much as pointing out the problem. I think the long term solution is to do the EOI after the interrupt is handled, even if it was deferred until doreti, but that probably requires assigning priorities to interrupts, and everything that goes along with that, maybe even making cpl a local-APIC priority instead of an IRQ mask. I'm not sure UNMASK_IRQ() should do the eoi either, since that can happen from doreti as well, right? Maybe the Xintrxx and Xresumexx should both call the handler, Xintrxx should do the EOI, and then join up for the doreti. ... but the problem with that is that lower priority interrupts are blocked for the duration of the handler ... so that seems to force the determination of interrupt priorities. It's funny that the APIC design seemingly forces the old kernel spl-priority model after BSD adapted to the PIC model. It seems to me the h/w and s/w designers both tried to adapt to the other and both ended up changing to be compatible and thereby ended up incompatible! It might take awhile before I'd be able to test any change in this area, I can't promise very quick turnaround. Joe From owner-freebsd-smp Wed Dec 3 18:31:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id SAA07404 for smp-outgoing; Wed, 3 Dec 1997 18:31:03 -0800 (PST) (envelope-from owner-freebsd-smp) 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 SAA07397 for ; Wed, 3 Dec 1997 18:30:58 -0800 (PST) (envelope-from smp@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id TAA10762; Wed, 3 Dec 1997 19:30:51 -0700 (MST) Message-Id: <199712040230.TAA10762@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Joe Eykholt Subject: Re: SMP Cc: smp@freebsd.org In-reply-to: Your message of "Wed, 03 Dec 1997 17:27:49 PST." <34860715.31DFF4F5@ipsilon.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 19:30:51 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Joe, > I'm glad you see the problem. I wasn't pressing for an immediate > solution so much as pointing out the problem. I think the long term nows the time, december is my "free month". --- > I'm not sure UNMASK_IRQ() should do the eoi either, since that can > happen from doreti as well, right? Maybe the Xintrxx and Xresumexx I actually mis-spoke last time, what I did was remove the lapic_eoi that was sent before the XResumeXXX entry point, and modify UNMASK_IRQ() to send it if the INT was currently UNmasked AND a level INT. The theory is that if its unmasked we must be in the first occurrance of an INT, ie NOT coming in from doreti(), as that could only occur for an INT that was defered, and thus would be masked. I applied the following patch to fix this: *** apic_vector.s 1997/12/04 01:11:26 1.37 --- apic_vector.s 1997/12/04 01:46:40 *************** *** 190,196 **** #define UNMASK_IRQ(irq_num) \ IMASK_LOCK ; /* into critical reg */ \ testl $IRQ_BIT(irq_num), _apic_imen ; \ ! je 9f ; \ andl $~IRQ_BIT(irq_num), _apic_imen ;/* clear mask bit */ \ movl _ioapic,%ecx ; /* ioapic[0]addr */ \ movl $REDTBL_IDX(irq_num),(%ecx) ; /* write the index */ \ --- 190,201 ---- #define UNMASK_IRQ(irq_num) \ IMASK_LOCK ; /* into critical reg */ \ testl $IRQ_BIT(irq_num), _apic_imen ; \ ! jne 7f ; /* bit set, masked */ \ ! testl $IRQ_BIT(irq_num), _apic_pin_trigger ; \ ! jz 9f ; /* edge, don't EOI */ \ ! movl $0, lapic_eoi ; /* should be safe */ \ ! jmp 9f ; /* skip unmasking */ \ ! 7: \ andl $~IRQ_BIT(irq_num), _apic_imen ;/* clear mask bit */ \ movl _ioapic,%ecx ; /* ioapic[0]addr */ \ movl $REDTBL_IDX(irq_num),(%ecx) ; /* write the index */ \ *************** *** 239,245 **** orl $IRQ_BIT(irq_num), _cil ; \ AVCPL_UNLOCK ; \ ; \ ! movl $0, lapic_eoi ; /* XXX too soon? */ \ incb _intr_nesting_level ; \ ; \ /* entry point used by doreti_unpend for HWIs. */ \ --- 244,250 ---- orl $IRQ_BIT(irq_num), _cil ; \ AVCPL_UNLOCK ; \ ; \ ! ;;; movl $0, lapic_eoi ; /* XXX too soon? */ \ incb _intr_nesting_level ; \ ; \ /* entry point used by doreti_unpend for HWIs. */ \ --- When I get a few reports I'll commit it. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Wed Dec 3 19:10:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id TAA10496 for smp-outgoing; Wed, 3 Dec 1997 19:10:22 -0800 (PST) (envelope-from owner-freebsd-smp) 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 TAA10486 for ; Wed, 3 Dec 1997 19:10:18 -0800 (PST) (envelope-from fbsd@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id UAA10940; Wed, 3 Dec 1997 20:10:11 -0700 (MST) Message-Id: <199712040310.UAA10940@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Steve Passe cc: Joe Eykholt , smp@freebsd.org Subject: Re: SMP In-reply-to: Your message of "Wed, 03 Dec 1997 19:30:51 MST." <199712040230.TAA10762@Ilsa.StevesCafe.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 03 Dec 1997 20:10:11 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > *** apic_vector.s 1997/12/04 01:11:26 1.37 > --- apic_vector.s 1997/12/04 01:46:40 > *************** > *** 239,245 **** > orl $IRQ_BIT(irq_num), _cil ; \ > AVCPL_UNLOCK ; \ > ; \ > ! movl $0, lapic_eoi ; /* XXX too soon? */ \ > incb _intr_nesting_level ; \ > ; \ > /* entry point used by doreti_unpend for HWIs. */ \ > --- 244,250 ---- > orl $IRQ_BIT(irq_num), _cil ; \ > AVCPL_UNLOCK ; \ > ; \ > ! ;;; movl $0, lapic_eoi ; /* XXX too soon? */ \ > incb _intr_nesting_level ; \ > ; \ > /* entry point used by doreti_unpend for HWIs. */ \ oops, the above is wrong, specifically the line with 3 ';' commenting out the lapic_eoi is the section not currently used. After applying this patch find the similar line in the 'else' section and comment that out also. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Thu Dec 4 03:55:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id DAA15002 for smp-outgoing; Thu, 4 Dec 1997 03:55:14 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.129.26]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id DAA14990 for ; Thu, 4 Dec 1997 03:55:03 -0800 (PST) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost.cybercity.dk [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.7) with ESMTP id MAA08362 for ; Thu, 4 Dec 1997 12:53:03 +0100 (CET) (envelope-from phk@critter.freebsd.dk) To: smp@freebsd.org Subject: PR400FX bios 1.08 is out... From: Poul-Henning Kamp Date: Thu, 04 Dec 1997 12:53:03 +0100 Message-ID: <8360.881236383@critter.freebsd.dk> Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk For those of you who have the Intel PR400FX ("Providence") motherboard: http://developer.intel.com/design/motherbd/pr/pr_bios.htm -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." From owner-freebsd-smp Thu Dec 4 07:24:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id HAA28718 for smp-outgoing; Thu, 4 Dec 1997 07:24:32 -0800 (PST) (envelope-from owner-freebsd-smp) 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 HAA28709 for ; Thu, 4 Dec 1997 07:24:27 -0800 (PST) (envelope-from jher@bermuda.io.com) Received: (from jher@localhost) by bermuda.io.com (8.8.5/8.8.5) id JAA14574 for freebsd-smp@freebsd.org; Thu, 4 Dec 1997 09:24:23 -0600 (CST) From: jher Message-Id: <199712041524.JAA14574@bermuda.io.com> Subject: Recent addition To: freebsd-smp@freebsd.org Date: Thu, 4 Dec 1997 09:24:22 -0600 (CST) X-Mailer: ELM [version 2.4//1 PL25] Content-Type: text Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, We recently brought up FreeBSD 3.0-971125-SNAP on an ALR Evolution Dual6 DT Motherboard (part # 74082101). Its working just fine so far and we've had marvelous results with it running innd (cnfs variant). We have 2 Pentium Pro 200, 9 stepping cpus in it and 64M of ram. I figured whomever is maintaining the list of "working MB's" might like this information. Thanks and have a nice day. From owner-freebsd-smp Thu Dec 4 10:49:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id KAA15526 for smp-outgoing; Thu, 4 Dec 1997 10:49:01 -0800 (PST) (envelope-from owner-freebsd-smp) 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 KAA15494; Thu, 4 Dec 1997 10:48:48 -0800 (PST) (envelope-from fbsd@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id LAA13864; Thu, 4 Dec 1997 11:48:45 -0700 (MST) Message-Id: <199712041848.LAA13864@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Poul-Henning Kamp cc: smp@freebsd.org Subject: Re: PR400FX bios 1.08 is out... In-reply-to: Your message of "Thu, 04 Dec 1997 12:53:03 +0100." <8360.881236383@critter.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 04 Dec 1997 11:48:45 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > For those of you who have the Intel PR400FX ("Providence") motherboard: > > http://developer.intel.com/design/motherbd/pr/pr_bios.htm does this have anything of note that would make one want to use this update? -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Thu Dec 4 10:55:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id KAA16298 for smp-outgoing; Thu, 4 Dec 1997 10:55:15 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.129.26]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id KAA16289 for ; Thu, 4 Dec 1997 10:55:05 -0800 (PST) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost.cybercity.dk [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.7) with ESMTP id TAA09167; Thu, 4 Dec 1997 19:52:55 +0100 (CET) (envelope-from phk@critter.freebsd.dk) To: Steve Passe cc: smp@freebsd.org Subject: Re: PR400FX bios 1.08 is out... In-reply-to: Your message of "Thu, 04 Dec 1997 11:48:45 MST." <199712041848.LAA13864@Ilsa.StevesCafe.com> Date: Thu, 04 Dec 1997 19:52:54 +0100 Message-ID: <9165.881261574@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In message <199712041848.LAA13864@Ilsa.StevesCafe.com>, Steve Passe writes: >Hi, > >> For those of you who have the Intel PR400FX ("Providence") motherboard: >> >> http://developer.intel.com/design/motherbd/pr/pr_bios.htm > >does this have anything of note that would make one want to use this update? There seems to be several interesting things in it, but I'm not sure yet. My StarOffice croaked on the .doc file :-( -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." From owner-freebsd-smp Thu Dec 4 12:23:50 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA25502 for smp-outgoing; Thu, 4 Dec 1997 12:23:50 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from romulus.hosteng.org (romulus.ihosteng.priv.no [195.0.205.14]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id MAA25483 for ; Thu, 4 Dec 1997 12:23:41 -0800 (PST) (envelope-from ivar@romulus.hosteng.org) Received: from localhost (ivar@localhost) by romulus.hosteng.org (8.8.8/8.8.7) with SMTP id VAA01099 for ; Thu, 4 Dec 1997 21:19:54 +0100 (CET) Date: Thu, 4 Dec 1997 21:19:54 +0100 (CET) From: Ivar Hosteng To: smp@freebsd.org Subject: FreeBSD 3.0-current SMP kernel gives a pagefault in kernel mode Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk After cvsup'ing FreeBSD 3.0-current and doing a make world followed by a kernel remake the new kernel refuses to boot. The cvsup was done thursday 4 at 13:00 GMT. The kenel gives a kernel panic stating a pagefault while in kernel mode after reporting real memory, but before reporting available memory. I am running a dual PPro 200Mhz system on a PR440FX motherboard. Regards, Ivar Hosteng email: ivar@hosteng.org From owner-freebsd-smp Thu Dec 4 13:08:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA01250 for smp-outgoing; Thu, 4 Dec 1997 13:08:21 -0800 (PST) (envelope-from owner-freebsd-smp) 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 NAA01227 for ; Thu, 4 Dec 1997 13:08:15 -0800 (PST) (envelope-from fbsd@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id OAA14422; Thu, 4 Dec 1997 14:08:05 -0700 (MST) Message-Id: <199712042108.OAA14422@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Ivar Hosteng cc: smp@freebsd.org Subject: Re: FreeBSD 3.0-current SMP kernel gives a pagefault in kernel mode In-reply-to: Your message of "Thu, 04 Dec 1997 21:19:54 +0100." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 04 Dec 1997 14:08:05 -0700 Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, > After cvsup'ing FreeBSD 3.0-current and doing a make world followed by a > kernel remake the new kernel refuses to boot. > > The cvsup was done thursday 4 at 13:00 GMT. > > The kenel gives a kernel panic stating a pagefault while in kernel mode > after reporting real memory, but before reporting available memory. > > I am running a dual PPro 200Mhz system on a PR440FX motherboard. I updated my world on: Dec 3 11:50 MST LOGS/standard-971203a With the exception of the gnu runtime lib, I had to make no changes. Am running the system now on exactly the same motherboard with no problems... -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-smp Thu Dec 4 16:39:11 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id QAA19434 for smp-outgoing; Thu, 4 Dec 1997 16:39:11 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from axis.axisnet.net (root@axis.axisnet.net [206.54.226.1]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id QAA19429 for ; Thu, 4 Dec 1997 16:39:05 -0800 (PST) (envelope-from jtk@titania.net) Received: from dega.titania.net (monet.titania.net [206.54.226.103]) by axis.axisnet.net (8.8.7/8.8.5) with SMTP id SAA14353 for ; Thu, 4 Dec 1997 18:39:54 -0600 Date: Thu, 4 Dec 97 18:28:53 Central Standard Time From: "Joseph T. Klein" Subject: "Sticky" Processes To: smp@freebsd.org X-Mailer: Chameleon ATX 6.0, Standards Based IntraNet Solutions, NetManage Inc. X-Priority: 3 (Normal) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Is it possible to give a process an attribute that would tend to force a processor into dedicating itself to it under SMP? A recent paper given at SigComm97 has revived my interest in using FreeBSD as a router platform. Check out: http://www.acm.org/sigcomm/sigcomm97/papers/p192.pdf It would be interesting to see how this can be implemented in an SMP environment. I essence holding the routing table lookup in cache on one processor during high demand. The implications of this paper are, to say the least, interesting. A hack of gated with these algorithms and SMP? -- From: Joseph T. Klein, Titania Corporation http://www.titania.net/ mailto: jtk@titania.net Sent: 18:28:54 CST/CDT 12/04/97 voice: +1 414 372 4565 FAX: +1 414 264 6038 From owner-freebsd-smp Thu Dec 4 17:25:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA23470 for smp-outgoing; Thu, 4 Dec 1997 17:25:01 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from icicle.winternet.com (adm@icicle.winternet.com [198.174.169.13]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id RAA23462 for ; Thu, 4 Dec 1997 17:24:57 -0800 (PST) (envelope-from mestery@mail.winternet.com) Received: (from adm@localhost) by icicle.winternet.com (8.8.8/8.8.8) id TAA24459; Thu, 4 Dec 1997 19:24:21 -0600 (CST) Received: from tundra.winternet.com(198.174.169.11) by icicle.winternet.com via smap (V2.0) id xma024289; Thu, 4 Dec 97 19:23:54 -0600 Received: from localhost (mestery@localhost) by tundra.winternet.com (8.8.7/8.8.4) with SMTP id TAA18928; Thu, 4 Dec 1997 19:23:54 -0600 (CST) X-Authentication-Warning: tundra.winternet.com: mestery owned process doing -bs Date: Thu, 4 Dec 1997 19:23:54 -0600 (CST) From: Kyle Mestery To: Steve Passe cc: Ivar Hosteng , smp@freebsd.org Subject: Re: FreeBSD 3.0-current SMP kernel gives a pagefault in kernel mode In-Reply-To: <199712042108.OAA14422@Ilsa.StevesCafe.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Thu, 4 Dec 1997, Steve Passe wrote: > Hi, > > > After cvsup'ing FreeBSD 3.0-current and doing a make world followed by a > > kernel remake the new kernel refuses to boot. > > > > The cvsup was done thursday 4 at 13:00 GMT. > > > > The kenel gives a kernel panic stating a pagefault while in kernel mode > > after reporting real memory, but before reporting available memory. > > > > I am running a dual PPro 200Mhz system on a PR440FX motherboard. > > I updated my world on: > Dec 3 11:50 MST LOGS/standard-971203a > > With the exception of the gnu runtime lib, I had to make no changes. > Am running the system now on exactly the same motherboard with no problems... > -- I updated my kernel at 6:30AM CST this morning. Both a UP and an SMP kernel page fault very early in the boot process. Kyle Mestery StorageTek's Network Systems Group 7600 Boone Ave. N., Minneapolis, MN 55428 mesteka@anubis.network.com, mestery@winternet.com "You do not greet Death, you punch him in the throat repeatedly until he drags you away." --No Fear From owner-freebsd-smp Thu Dec 4 17:42:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id RAA24776 for smp-outgoing; Thu, 4 Dec 1997 17:42:42 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from zed.ludd.luth.se (zed.ludd.luth.se [130.240.16.33]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id RAA24769 for ; Thu, 4 Dec 1997 17:42:35 -0800 (PST) (envelope-from pantzer@brother.ludd.luth.se) Received: from brother.ludd.luth.se (pantzer@brother.ludd.luth.se [130.240.16.78]) by zed.ludd.luth.se (8.8.5/8.8.5) with SMTP id CAA27333; Fri, 5 Dec 1997 02:42:22 +0100 Message-Id: <199712050142.CAA27333@zed.ludd.luth.se> X-Mailer: exmh version 2.0zeta 7/24/97 To: "Joseph T. Klein" cc: smp@freebsd.org Subject: Re: "Sticky" Processes In-reply-to: Your message of "Thu, 04 Dec 1997 18:28:53 EST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 05 Dec 1997 02:42:18 +0100 From: Mattias Pantzare Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > It would be interesting to see how this can be implemented in > an SMP environment. I essence holding the routing table lookup in > cache on one processor during high demand. The way that they did a test implementation of it was to hack NetBSD (no SMP support in that NetBSD) and hacked it to do the routing processing on the second CPU. All packets that needen more processing was forwarded to the CPU that run NetBSD. That way the routing table was allways in the CPU cache (except for when it was updated). (I don't remember any performance figures now, but I can ask if anyone is interested) From owner-freebsd-smp Thu Dec 4 22:03:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id WAA18793 for smp-outgoing; Thu, 4 Dec 1997 22:03:25 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA18788 for ; Thu, 4 Dec 1997 22:03:15 -0800 (PST) (envelope-from bde@zeta.org.au) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.7/8.6.9) id QAA19906; Fri, 5 Dec 1997 16:58:40 +1100 Date: Fri, 5 Dec 1997 16:58:40 +1100 From: Bruce Evans Message-Id: <199712050558.QAA19906@godzilla.zeta.org.au> To: ivar@romulus.hosteng.org, smp@freebsd.org Subject: Re: FreeBSD 3.0-current SMP kernel gives a pagefault in kernel mode Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >After cvsup'ing FreeBSD 3.0-current and doing a make world followed by a >kernel remake the new kernel refuses to boot. Malloc initialization was broken yesterday. This causes a measly 64 compiler wanings for GENERIC and an early panic. (The malloc cookie is too corrupt for the magic number check to be reached (ks_next initially points to a string so it is non-NULL. malloc() somehow avoids crashing an returns NULL. Some early callers actually check for the NULL return, but are remarkably silent about the error. Here, the first caller that doesn't check is in cpu_startup(), and a panic occurs when the NULL pointer is bzeroed.) Bruce diff -c2 malloc.h~ malloc.h *** malloc.h~ Fri Dec 5 03:56:37 1997 --- malloc.h Fri Dec 5 04:37:05 1997 *************** *** 64,68 **** #define MALLOC_DEFINE(type, shortdesc, longdesc) \ ! struct malloc_type type[1] = { { shortdesc, M_MAGIC } }; \ struct __hack --- 64,70 ---- #define MALLOC_DEFINE(type, shortdesc, longdesc) \ ! struct malloc_type type[1] = { \ ! { NULL, 0, 0, 0, 0, 0, 0, M_MAGIC, shortdesc, 0, 0 } \ ! }; \ struct __hack From owner-freebsd-smp Thu Dec 4 22:26:46 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id WAA21182 for smp-outgoing; Thu, 4 Dec 1997 22:26:46 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id WAA21169 for ; Thu, 4 Dec 1997 22:26:40 -0800 (PST) (envelope-from toor@dyson.iquest.net) Received: (from root@localhost) by dyson.iquest.net (8.8.7/8.8.8) id BAA00319; Fri, 5 Dec 1997 01:26:18 -0500 (EST) (envelope-from toor) From: "John S. Dyson" Message-Id: <199712050626.BAA00319@dyson.iquest.net> Subject: Re: FreeBSD 3.0-current SMP kernel gives a pagefault in kernel mode In-Reply-To: <199712050558.QAA19906@godzilla.zeta.org.au> from Bruce Evans at "Dec 5, 97 04:58:40 pm" To: bde@zeta.org.au (Bruce Evans) Date: Fri, 5 Dec 1997 01:26:18 -0500 (EST) Cc: ivar@romulus.hosteng.org, smp@freebsd.org X-Mailer: ELM [version 2.4ME+ PL31 (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 Bruce Evans said: > >After cvsup'ing FreeBSD 3.0-current and doing a make world followed by a > >kernel remake the new kernel refuses to boot. > > Malloc initialization was broken yesterday. This causes a measly 64 > compiler wanings for GENERIC and an early panic. (The malloc cookie > is too corrupt for the magic number check to be reached (ks_next > initially points to a string so it is non-NULL. malloc() somehow > avoids crashing an returns NULL. Some early callers actually check > for the NULL return, but are remarkably silent about the error. Here, > the first caller that doesn't check is in cpu_startup(), and a panic > occurs when the NULL pointer is bzeroed.) > I committed the fix to current (thanks to John Hood.) -- John dyson@freebsd.org jdyson@nc.com From owner-freebsd-smp Thu Dec 4 23:49:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id XAA28424 for smp-outgoing; Thu, 4 Dec 1997 23:49:01 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from verdi.nethelp.no (verdi.nethelp.no [195.1.171.130]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id XAA28393 for ; Thu, 4 Dec 1997 23:48:43 -0800 (PST) (envelope-from sthaug@nethelp.no) From: sthaug@nethelp.no Received: (qmail 8566 invoked by uid 1001); 5 Dec 1997 07:41:27 +0000 (GMT) To: jtk@titania.net Cc: smp@freebsd.org Subject: Re: "Sticky" Processes In-Reply-To: Your message of "Thu, 4 Dec 97 18:28:53 Central Standard Time" References: X-Mailer: Mew version 1.05+ on Emacs 19.28.2 Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Date: Fri, 05 Dec 1997 08:41:27 +0100 Message-ID: <8564.881307687@verdi.nethelp.no> Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > A recent paper given at SigComm97 has revived my interest in using > FreeBSD as a router platform. > > Check out: > > http://www.acm.org/sigcomm/sigcomm97/papers/p192.pdf Ah yes, the Degermark et al. paper. It's already become rather well known in networking circles. > It would be interesting to see how this can be implemented in > an SMP environment. I essence holding the routing table lookup in > cache on one processor during high demand. > > The implications of this paper are, to say the least, interesting. You'd need to do a *lot* of work on the FreeBSD TCP/IP stack to be able to *use* two million forwarding decisions per second, though. Steinar Haug, Nethelp consulting, sthaug@nethelp.no From owner-freebsd-smp Fri Dec 5 12:42:37 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id MAA06266 for smp-outgoing; Fri, 5 Dec 1997 12:42:37 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id MAA06260 for ; Fri, 5 Dec 1997 12:42:34 -0800 (PST) (envelope-from dcmyers@concord.Corp.Sun.COM) Received: from Corp.Sun.COM ([129.145.35.78]) by mercury.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id MAA04002 for ; Fri, 5 Dec 1997 12:41:57 -0800 Received: from mpk05.Corp.Sun.COM by Corp.Sun.COM (SMI-8.6/SMI-5.3) id MAA20423; Fri, 5 Dec 1997 12:41:53 -0800 Received: from concord.Corp.Sun.COM by mpk05.Corp.Sun.COM (SMI-8.6/SMI-SVR4) id MAA20982; Fri, 5 Dec 1997 12:41:55 -0800 Received: from concord.corp.sun.com by concord.Corp.Sun.COM (SMI-8.6/SMI-SVR4) id MAA21802; Fri, 5 Dec 1997 12:41:53 -0800 Message-Id: <199712052041.MAA21802@concord.Corp.Sun.COM> Date: Fri, 5 Dec 1997 12:41:50 -0800 (PST) From: david.myers@Corp.Sun.COM Subject: A pointed question: SMP directions... To: freebsd-smp@freebsd.org MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Folks: Let me ask a pointed question as... delicately... as I can. I've been using FreeBSD on my home machine for close to two years. Like it a lot. Gotten very comfortable with it. Particularly like the ports collection and the ease of bringing software up on it. But I've got this nifty dual Pentium Pro box. As it happens, I also work for Sun, doing mostly Java work. So Solaris 2.6 looms large. The pointed question to this list: rate yourself, guys. Rate your progress in getting multiprocessor/multithreading BSD up and running. How close are you to matching the SMP features of the Solaris kernel? (And I think back to the couple of bad years we had when our kernel was having growing pains...) Now, add to that the difficulties in getting Java up and running under FreeBSD -- Kaffe will be very nice *someday* -- and what can I say? I'm leaning towards Solaris. Note that I'm not a kernel hacker, and 90% solutions are probably fine. I'm looking for that warm, fuzzy feeling that SMP FreeBSD will be "rock solid", as the CD packaging says. I'm just thinking, we've got some awfully smart people here at Sun, and it sure took them a long time to get SMP right. Now, of course, Solaris is a little more piggish than BSD -- "meant for mission-critical applications" as we would say around the office. I've never liked SysV from a user and administrator point of view, but I've learned to cope, what with a Sparc box sitting on my desk. And attempting to master Solaris' bizarre PPP implementation is always a joy -- a real problem for a home/hacker machine. But that out-of-the-box multi-CPU support and the kernel Java support look nice... So I'd love to hear what the roadmap looks like; what the projected feature set of SMP FreeBSD will look like. And timelines. How far do you think you'll get by the time 3.0-RELEASE comes out? Comments from users making/contemplating similar moves will also be appreciated. -David. From owner-freebsd-smp Fri Dec 5 13:18:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id NAA09903 for smp-outgoing; Fri, 5 Dec 1997 13:18:20 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id NAA09897 for ; Fri, 5 Dec 1997 13:18:16 -0800 (PST) (envelope-from mrcpu@cdsnet.net) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by mail.cdsnet.net (8.8.6/8.8.6) with SMTP id NAA24899; Fri, 5 Dec 1997 13:11:39 -0800 (PST) Date: Fri, 5 Dec 1997 13:11:37 -0800 (PST) From: Jaye Mathisen To: david.myers@Corp.Sun.COM cc: freebsd-smp@freebsd.org Subject: Re: A pointed question: SMP directions... In-Reply-To: <199712052041.MAA21802@concord.Corp.Sun.COM> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I think the instability in -current is greatly overrated. I have just taken the path of "When I find a snapshot that works, hell must chill before I move to a new snapshot". Of course, the more people that install it and use it, and report bugs, the sooner it will be more stable, so there's kind of a chicken and egg problem. My 3.0 boxes from around the beginning of Oct (10/8) have been just solid little clunkers. I would suggest you install it and try it. You can always delete it. Of course, I can only speak as a pretty satisfied user with something like 4 3.0 boxes (all duals, the quad's firing up next week). Now the multithreading issue is a whole 'nudder story, but since I currently don't have a need for it, I haven't been following the discussions too closely, but IIRC, kernel threading was a way off, and I think was hampered a bit with some of the locking issues in the kernel that were being discussed recently. But I could be mixing apples and kiwis. On Fri, 5 Dec 1997 david.myers@Corp.Sun.COM wrote: > > Folks: > > Let me ask a pointed question as... delicately... as I can. I've been > using FreeBSD on my home machine for close to two years. Like it a > lot. Gotten very comfortable with it. Particularly like the ports > collection and the ease of bringing software up on it. But I've got > this nifty dual Pentium Pro box. As it happens, I also work for Sun, > doing mostly Java work. So Solaris 2.6 looms large. > > The pointed question to this list: rate yourself, guys. Rate your > progress in getting multiprocessor/multithreading BSD up and running. > How close are you to matching the SMP features of the Solaris kernel? > (And I think back to the couple of bad years we had when our kernel was > having growing pains...) Now, add to that the difficulties in getting > Java up and running under FreeBSD -- Kaffe will be very nice *someday* > -- and what can I say? I'm leaning towards Solaris. > > Note that I'm not a kernel hacker, and 90% solutions are probably > fine. I'm looking for that warm, fuzzy feeling that SMP FreeBSD will > be "rock solid", as the CD packaging says. I'm just thinking, we've > got some awfully smart people here at Sun, and it sure took them a long > time to get SMP right. > > Now, of course, Solaris is a little more piggish than BSD -- "meant for > mission-critical applications" as we would say around the office. I've > never liked SysV from a user and administrator point of view, but I've > learned to cope, what with a Sparc box sitting on my desk. And > attempting to master Solaris' bizarre PPP implementation is always a joy > -- a real problem for a home/hacker machine. But that out-of-the-box > multi-CPU support and the kernel Java support look nice... > > So I'd love to hear what the roadmap looks like; what the projected > feature set of SMP FreeBSD will look like. And timelines. How > far do you think you'll get by the time 3.0-RELEASE comes out? Comments > from users making/contemplating similar moves will also be appreciated. > > -David. > > From owner-freebsd-smp Fri Dec 5 15:22:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA22302 for smp-outgoing; Fri, 5 Dec 1997 15:22:34 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from send1a.yahoomail.com (send1a.yahoomail.com [205.180.60.22]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id PAA22283 for ; Fri, 5 Dec 1997 15:22:27 -0800 (PST) (envelope-from rgireyev@yahoo.com) Message-ID: <19971205232212.1671.rocketmail@send1a.yahoomail.com> Received: from [156.153.255.234] by send1a; Fri, 05 Dec 1997 15:22:12 PST Date: Fri, 5 Dec 1997 15:22:12 -0800 (PST) From: Rudy Gireyev Subject: Re: A pointed question: SMP directions... To: Jaye Mathisen , david.myers@Corp.Sun.COM Cc: freebsd-smp@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I didn't realize that kernel level JAVA support was even being contemplated. Jaye what do you mean by saying you currently have no need for kernel threading support????? Lastly, David, if you want to experiment you can dual boot the machine with FreeBSD and Solaris. Good way to see first hand the performance differences :-) ---Jaye Mathisen wrote: > > > I think the instability in -current is greatly overrated. I have just > taken the path of "When I find a snapshot that works, hell must chill > before I move to a new snapshot". > > Of course, the more people that install it and use it, and report bugs, > the sooner it will be more stable, so there's kind of a chicken and egg > problem. > > My 3.0 boxes from around the beginning of Oct (10/8) have been just > solid little clunkers. > > I would suggest you install it and try it. You can always delete it. > > Of course, I can only speak as a pretty satisfied user with something like > 4 3.0 boxes (all duals, the quad's firing up next week). > > Now the multithreading issue is a whole 'nudder story, but since I > currently don't have a need for it, I haven't been following the > discussions too closely, but IIRC, kernel threading was a way off, and > I think was hampered a bit with some of the locking issues in the kernel > that were being discussed recently. But I could be mixing apples and > kiwis. > > On Fri, 5 Dec 1997 david.myers@Corp.Sun.COM wrote: > > > > > Folks: > > > > Let me ask a pointed question as... delicately... as I can. I've been > > using FreeBSD on my home machine for close to two years. Like it a > > lot. Gotten very comfortable with it. Particularly like the ports > > collection and the ease of bringing software up on it. But I've got > > this nifty dual Pentium Pro box. As it happens, I also work for Sun, > > doing mostly Java work. So Solaris 2.6 looms large. > > > > The pointed question to this list: rate yourself, guys. Rate your > > progress in getting multiprocessor/multithreading BSD up and running. > > How close are you to matching the SMP features of the Solaris kernel? > > (And I think back to the couple of bad years we had when our kernel was > > having growing pains...) Now, add to that the difficulties in getting > > Java up and running under FreeBSD -- Kaffe will be very nice *someday* > > -- and what can I say? I'm leaning towards Solaris. > > > > Note that I'm not a kernel hacker, and 90% solutions are probably > > fine. I'm looking for that warm, fuzzy feeling that SMP FreeBSD will > > be "rock solid", as the CD packaging says. I'm just thinking, we've > > got some awfully smart people here at Sun, and it sure took them a long > > time to get SMP right. > > > > Now, of course, Solaris is a little more piggish than BSD -- "meant for > > mission-critical applications" as we would say around the office. I've > > never liked SysV from a user and administrator point of view, but I've > > learned to cope, what with a Sparc box sitting on my desk. And > > attempting to master Solaris' bizarre PPP implementation is always a joy > > -- a real problem for a home/hacker machine. But that out-of-the-box > > multi-CPU support and the kernel Java support look nice... > > > > So I'd love to hear what the roadmap looks like; what the projected > > feature set of SMP FreeBSD will look like. And timelines. How > > far do you think you'll get by the time 3.0-RELEASE comes out? Comments > > from users making/contemplating similar moves will also be appreciated. > > > > -David. > > > > > > _________________________________________________________ DO YOU YAHOO!? Get your free @yahoo.com address at http://mail.yahoo.com From owner-freebsd-smp Fri Dec 5 15:35:58 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA23557 for smp-outgoing; Fri, 5 Dec 1997 15:35:58 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id PAA23550 for ; Fri, 5 Dec 1997 15:35:55 -0800 (PST) (envelope-from dcmyers@concord.Corp.Sun.COM) Received: from Corp.Sun.COM ([129.145.35.78]) by mercury.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id PAA17078; Fri, 5 Dec 1997 15:35:24 -0800 Received: from mpk05.Corp.Sun.COM by Corp.Sun.COM (SMI-8.6/SMI-5.3) id PAA14890; Fri, 5 Dec 1997 15:35:21 -0800 Received: from concord.Corp.Sun.COM by mpk05.Corp.Sun.COM (SMI-8.6/SMI-SVR4) id PAA24820; Fri, 5 Dec 1997 15:35:21 -0800 Received: from concord.corp.sun.com by concord.Corp.Sun.COM (SMI-8.6/SMI-SVR4) id PAA23091; Fri, 5 Dec 1997 15:35:20 -0800 Message-Id: <199712052335.PAA23091@concord.Corp.Sun.COM> Date: Fri, 5 Dec 1997 15:35:16 -0800 (PST) From: david.myers@Corp.Sun.COM Subject: Re: A pointed question: SMP directions... To: rgireyev@yahoo.com cc: freebsd-smp@freebsd.org In-Reply-To: <19971205232212.1671.rocketmail@send1a.yahoomail.com> MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > I didn't realize that kernel level JAVA support was even being > contemplated. Under Solaris, not FreeBSD, so far as I know. > Lastly, David, if you want to experiment you can dual boot the machine > with FreeBSD and Solaris. Good way to see first hand the performance > differences :-) Perhaps I'm not being clear. I "experiment" (if you can call it that) with Solaris every day at work. That's not what I'm asking about. I want to know how much progress the SMP team in FreeBSD thinks it's making. I want to know where they're headed, what problems they see; what they'd wish for in the 3.0 release. I want to know how long it will take before FreeBSD matches (or comes close to matching) the fine-grained SMP support that Solaris has. In short, I'm looking for both technical and emotional reasons for staying with FreeBSD. I like FreeBSD and would like to stay with it, but the feature list of Solaris relative to my applications (i.e., Java) and hardware (i.e., dual PPro box) looks compelling. -David. ^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ David Myers Sun Microsystems Java Technologist david.myers@corp.sun.com Strategic Sales Area (650) 614-6907 ^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^v^ From owner-freebsd-smp Fri Dec 5 15:37:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA23693 for smp-outgoing; Fri, 5 Dec 1997 15:37:15 -0800 (PST) (envelope-from owner-freebsd-smp) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id PAA23688 for ; Fri, 5 Dec 1997 15:37:12 -0800 (PST) (envelope-from mrcpu@cdsnet.net) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by mail.cdsnet.net (8.8.6/8.8.6) with SMTP id PAA23611; Fri, 5 Dec 1997 15:36:46 -0800 (PST) Date: Fri, 5 Dec 1997 15:36:46 -0800 (PST) From: Jaye Mathisen To: Rudy Gireyev cc: david.myers@Corp.Sun.COM, freebsd-smp@freebsd.org Subject: Re: A pointed question: SMP directions... In-Reply-To: <19971205232212.1671.rocketmail@send1a.yahoomail.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Fri, 5 Dec 1997, Rudy Gireyev wrote: > I didn't realize that kernel level JAVA support was even being > contemplated. I didn't mean to imply this, I meant kernel threads, which doesn't mean Java necessarily. > > Jaye what do you mean by saying you currently have no need for kernel > threading support????? None of my software that I need for what I do is non-functional due to a lack of kernel threads...