From owner-freebsd-smp Sun Jun 28 03:44:40 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id DAA02369 for freebsd-smp-outgoing; Sun, 28 Jun 1998 03:44:40 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id DAA02178 for ; Sun, 28 Jun 1998 03:42:55 -0700 (PDT) (envelope-from jhay@zibbi.mikom.csir.co.za) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.9.0/8.9.0) id MAA04522 for freebsd-smp@freebsd.org; Sun, 28 Jun 1998 12:42:31 +0200 (SAT) From: John Hay Message-Id: <199806281042.MAA04522@zibbi.mikom.csir.co.za> Subject: time problem? To: freebsd-smp@FreeBSD.ORG Date: Sun, 28 Jun 1998 12:42:31 +0200 (SAT) X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, I have found that my dual P5 100MHz machine will gain +- 0.85 seconds every now and again. I haven't been able to find out if there is a pattern yet. It seems to happen within a second or so, so it isn't that it slowly drifts. It happens on average a few times per day, but I have seen days where it hasn't happened at all. I first started to notice it in ntpd's logs, but I have since stopped using ntpd, to make sure that it isn't something that was induced by ntpd. I have also tried a non SMP kernel (with a hack to force it to also use 8254 for its time) and it does not seem to happen then, so it looks like it has something to do with the SMP code. It also does not seem to have anything to do with how busy the machine is, it has happened on a totally idle machine and also during a make world. Anybody have any ideas how I can look for this? I don't understand all the low level time stuff yet (especially the SMP side of it), but if given a little direction I'm willing to try a few things. John -- John Hay -- John.Hay@mikom.csir.co.za To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Jun 28 04:51:47 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id EAA11216 for freebsd-smp-outgoing; Sun, 28 Jun 1998 04:51:47 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.133.1] (may be forged)) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id EAA11112 for ; Sun, 28 Jun 1998 04:49:54 -0700 (PDT) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.5) with ESMTP id NAA02722; Sun, 28 Jun 1998 13:40:53 +0200 (CEST) To: John Hay cc: freebsd-smp@FreeBSD.ORG Subject: Re: time problem? In-reply-to: Your message of "Sat, 28 Jun 1998 12:42:31 +0200." <199806281042.MAA04522@zibbi.mikom.csir.co.za> Date: Sun, 28 Jun 1998 13:40:51 +0200 Message-ID: <2720.899034051@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >I have found that my dual P5 100MHz machine will gain +- 0.85 seconds >every now and again. [...] >so it looks like it has something >to do with the SMP code. It also does not seem to have anything to do >with how busy the machine is, it has happened on a totally idle machine >and also during a make world. > >Anybody have any ideas how I can look for this? I don't understand all >the low level time stuff yet (especially the SMP side of it), but if >given a little direction I'm willing to try a few things. It is really very simple, and there is nothing SMP specific about the code. .85 seconds sounds suspiciosly like 2^20 cycles at 1193182 Hz. Unfortunately that doesn't really make any kind of sense to me... How do you detect this jump ? what is the exact sequnce of events ? The only possibly weak spot I know is this test in i386/isa/clock.c, you can try out this patch, but I doubt it will fix it. Index: clock.c =================================================================== RCS file: /home/ncvs/src/sys/i386/isa/clock.c,v retrieving revision 1.124 diff -u -r1.124 clock.c --- clock.c 1998/06/09 13:10:46 1.124 +++ clock.c 1998/06/28 11:39:03 @@ -1148,7 +1148,7 @@ high = inb(TIMER_CNTR0); count = hardclock_max_count - ((high << 8) | low); - if (count < i8254_lastcount) { + if (count <= i8254_lastcount) { i8254_ticked = 1; i8254_offset += hardclock_max_count; } -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." "ttyv0" -- What UNIX calls a $20K state-of-the-art, 3D, hi-res color terminal To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Jun 28 11:25:42 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA21983 for freebsd-smp-outgoing; Sun, 28 Jun 1998 11:25:42 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-12.waterford.indigo.ie [194.125.139.75]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA21971 for ; Sun, 28 Jun 1998 11:25:33 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id TAA01235 for freebsd-smp@freebsd.org; Sun, 28 Jun 1998 19:21:41 +0100 (IST) (envelope-from rotel@indigo.ie) From: Niall Smart Message-Id: <199806281821.TAA01235@indigo.ie> Date: Sun, 28 Jun 1998 19:21:41 +0000 Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: freebsd-smp@FreeBSD.ORG Subject: PPro vs PII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Hi, I'm trying to decide between a system based on 2x PPro 200Mhz w/256KB cache or 2x PII 300Mhz. Does anyone have any relevant benchmarks? Niall -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org Annoy your enemies and astonish your friends: echo "#define if(x) if (!(x))" >> /usr/include/stdio.h To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Jun 28 18:57:10 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA15508 for freebsd-smp-outgoing; Sun, 28 Jun 1998 18:57:10 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA15498 for ; Sun, 28 Jun 1998 18:57:00 -0700 (PDT) (envelope-from jhay@zibbi.mikom.csir.co.za) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.9.0/8.9.0) id UAA09849; Sun, 28 Jun 1998 20:14:48 +0200 (SAT) From: John Hay Message-Id: <199806281814.UAA09849@zibbi.mikom.csir.co.za> Subject: Re: time problem? In-Reply-To: <2720.899034051@critter.freebsd.dk> from Poul-Henning Kamp at "Jun 28, 98 01:40:51 pm" To: phk@critter.freebsd.dk (Poul-Henning Kamp) Date: Sun, 28 Jun 1998 20:14:48 +0200 (SAT) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >I have found that my dual P5 100MHz machine will gain +- 0.85 seconds > >every now and again. [...] > >so it looks like it has something > >to do with the SMP code. It also does not seem to have anything to do > >with how busy the machine is, it has happened on a totally idle machine > >and also during a make world. > > > >Anybody have any ideas how I can look for this? I don't understand all > >the low level time stuff yet (especially the SMP side of it), but if > >given a little direction I'm willing to try a few things. > > It is really very simple, and there is nothing SMP specific about the > code. > > .85 seconds sounds suspiciosly like 2^20 cycles at 1193182 Hz. Ok, it is possible that I just haven't waited long enough with the UP kernel, but I can try again. > > Unfortunately that doesn't really make any kind of sense to me... > > How do you detect this jump ? what is the exact sequnce of events ? Well, it is actually the one that you gave me to test the new pps api with, which I modified a little. I'll attach it at the end. I have an 1 pps signal connected to DCD of my serial port. With sysctl kern.timecounter.frequency I tune the frequency until it runs as accurately as possible. Then I use "ntpdate -b " a few times to get the offset small. Then I run my little program and wait until it shows a jump. Then I use "ntpdate -d " to make sure that it was really a jump. Basically what the program does is to fetch the timestamp from the kernel and if it is within bounds, printf("...\r") it. If it is out of bounds, it add a printf("\n") before and after it. The values that it print is the timestamp (sec.nanosec), the sequence number and the time difference between this and the previous timestamp/. Here is the output of a piece of it: ---------------- A: 899000790.981073054 #29011 D: 1.000004191 A: 899000792.826104370 #29012 D: 1.845031316 A: 899016109.813482765 #44329 D: 0.999940488 A: 899016111.668519664 #44330 D: 1.855036899 ---------- So on sequence numbers 29012 and 44330, I have gained .85 seconds in the space of a second. For the 15000 seconds between 29012 and 44329 there were no glitches or jumps. > > The only possibly weak spot I know is this test in i386/isa/clock.c, > you can try out this patch, but I doubt it will fix it. Ok, I'll try it. I do also get glitches where the time will jump forward or backward by 1 tick (10ms), but it will correct itself on the next sample, so it does not have a long term effect. I thought its source was also this piece of code. I get about 30-50 of these per day, using my test program. Is this function, i8254_get_timecount(), used by something that can have a permanent effect on time? I thought it was only used by nanotime() and microtime() to get the offset from the previous tick, but that it wasn't used to calculate the next tick? > > Index: clock.c > =================================================================== > RCS file: /home/ncvs/src/sys/i386/isa/clock.c,v > retrieving revision 1.124 > diff -u -r1.124 clock.c > --- clock.c 1998/06/09 13:10:46 1.124 > +++ clock.c 1998/06/28 11:39:03 > @@ -1148,7 +1148,7 @@ > high = inb(TIMER_CNTR0); > > count = hardclock_max_count - ((high << 8) | low); > - if (count < i8254_lastcount) { > + if (count <= i8254_lastcount) { > i8254_ticked = 1; > i8254_offset += hardclock_max_count; > } John -- John Hay -- John.Hay@mikom.csir.co.za ----------------------------- ppstst2.c ----------------------------- #include #include #include #include #include #include #include #if 1 #define DEVNAME "/dev/gps0" #else #define DEVNAME "/dev/lppps0" #endif int main(int argc, char **argv) { int fd; struct timespec dts; pps_info_t pi, opi; pps_params_t pp; pps_handle_t ph; int i, mode; float df; fd = open(DEVNAME, O_RDONLY); if (fd < 0) err(1, DEVNAME); i = time_pps_create(fd, &ph); if (i < 0) err(1, "time_pps_create"); i = time_pps_getcap(ph, &mode); if (i < 0) err(1, "time_pps_getcap"); printf("getcap says %x\n", mode); pp.mode = PPS_CAPTUREASSERT; i = time_pps_setparams(ph, &pp); if (i < 0) err(1, "time_pps_setparams"); i = time_pps_fetch(ph, &pi); if (i < 0) err(1, "time_pps_fetch"); opi = pi; while (1) { i = time_pps_fetch(ph, &pi); if (i < 0) err(1, "time_pps_fetch"); if (opi.assert_sequence != pi.assert_sequence) { dts.tv_sec = pi.assert_timestamp.tv_sec - opi.assert_timestamp.tv_sec; dts.tv_nsec = pi.assert_timestamp.tv_nsec - opi.assert_timestamp.tv_nsec; if (dts.tv_nsec < 0) { dts.tv_sec--; dts.tv_nsec += 1000000000; } df = (float)dts.tv_sec; df += (float)dts.tv_nsec / 1000000000.0; if (df > 1.001 || df < 0.999) printf("\n"); printf("A: %d.%09d #%d D: %d.%09d\r", pi.assert_timestamp.tv_sec, pi.assert_timestamp.tv_nsec, pi.assert_sequence, dts.tv_sec, dts.tv_nsec); if (df > 1.001 || df < 0.999) printf("\n"); else fflush(NULL); opi = pi; } #if 0 printf("A: %d.%09d #%d C: %d.%09d #%d\n", pi.assert_timestamp.tv_sec, pi.assert_timestamp.tv_nsec, pi.assert_sequence, pi.clear_timestamp.tv_sec, pi.clear_timestamp.tv_nsec, pi.clear_sequence); #endif usleep(250000); } return(0); } To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Jun 28 20:42:02 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA28966 for freebsd-smp-outgoing; Sun, 28 Jun 1998 20:42:02 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id UAA28943 for ; Sun, 28 Jun 1998 20:41:57 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 5669 invoked by uid 1017); 29 Jun 1998 02:38:52 -0000 Date: Sun, 28 Jun 1998 20:38:52 -0600 (MDT) From: Atipa To: Niall Smart cc: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: <199806281821.TAA01235@indigo.ie> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can utilize 100MHz system bus. Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock rates, and much faster DRAM access, you'll definitely notice the difference. Pros are at the end of their lifecycle, and will be hard to support. Kevin On Sun, 28 Jun 1998, Niall Smart wrote: > Hi, > > I'm trying to decide between a system based on 2x PPro 200Mhz > w/256KB cache or 2x PII 300Mhz. Does anyone have any relevant > benchmarks? > > Niall > > > -- > Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk > FreeBSD: Turning PC's into Workstations: www.freebsd.org > Annoy your enemies and astonish your friends: > echo "#define if(x) if (!(x))" >> /usr/include/stdio.h > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-smp" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Sun Jun 28 23:59:01 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA18728 for freebsd-smp-outgoing; Sun, 28 Jun 1998 23:59:01 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.133.1] (may be forged)) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA18690 for ; Sun, 28 Jun 1998 23:58:51 -0700 (PDT) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.5) with ESMTP id IAA09572; Mon, 29 Jun 1998 08:56:27 +0200 (CEST) To: John Hay cc: freebsd-smp@FreeBSD.ORG Subject: Re: time problem? In-reply-to: Your message of "Sat, 28 Jun 1998 20:14:48 +0200." <199806281814.UAA09849@zibbi.mikom.csir.co.za> Date: Mon, 29 Jun 1998 08:56:26 +0200 Message-ID: <9570.899103386@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In message <199806281814.UAA09849@zibbi.mikom.csir.co.za>, John Hay writes: >I have an 1 pps signal connected to DCD of my serial port. What source do you use for this ? >---------------- >A: 899000790.981073054 #29011 D: 1.000004191 >A: 899000792.826104370 #29012 D: 1.845031316 >A: 899016109.813482765 #44329 D: 0.999940488 >A: 899016111.668519664 #44330 D: 1.855036899 >---------- > >So on sequence numbers 29012 and 44330, I have gained .85 seconds >in the space of a second. For the 15000 seconds between 29012 and >44329 there were no glitches or jumps. Now, one source of this could be interrupts diasabled for too long, since your timestamps depend on the sio interrupt that would be my primary suspect. Is there anything that makes it happen more often ? >Is this function, i8254_get_timecount(), used by something that can >have a permanent effect on time? I thought it was only used by >nanotime() and microtime() to get the offset from the previous tick, >but that it wasn't used to calculate the next tick? It shouldn't be, apart from "tco_forward()" which bumps the timecounters reference time once each hardclock() (kern/kern_clock.c) -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." "ttyv0" -- What UNIX calls a $20K state-of-the-art, 3D, hi-res color terminal To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 02:05:21 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA06751 for freebsd-smp-outgoing; Mon, 29 Jun 1998 02:05:21 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: (from sos@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA06728; Mon, 29 Jun 1998 02:05:11 -0700 (PDT) (envelope-from sos) Message-Id: <199806290905.CAA06728@hub.freebsd.org> Subject: Re: PPro vs PII In-Reply-To: from Atipa at "Jun 28, 98 08:38:52 pm" To: freebsd@atipa.com (Atipa) Date: Mon, 29 Jun 1998 02:05:09 -0700 (PDT) Cc: rotel@indigo.ie, freebsd-smp@FreeBSD.ORG From: sos@FreeBSD.ORG Reply-to: sos@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In reply to Atipa who wrote: > > The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > utilize 100MHz system bus. > Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock > rates, and much faster DRAM access, you'll definitely notice the > difference. Pros are at the end of their lifecycle, and will be hard to > support. Yeah, but the P6 runs its cache at full CPU speed the P2 doesn't, that hurts the P2 in SMP systems. In fact the 2x300P2 I've seen compares badly to my 2x200P6 system, and they cost alot more. P6 systems are a real bargain at the moment, as long as stocks last... -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Soren Schmidt (sos@FreeBSD.org) FreeBSD Core Team So much code to hack -- so little time. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 02:51:38 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id CAA12625 for freebsd-smp-outgoing; Mon, 29 Jun 1998 02:51:38 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id CAA12549 for ; Mon, 29 Jun 1998 02:51:18 -0700 (PDT) (envelope-from jhay@zibbi.mikom.csir.co.za) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.9.0/8.9.0) id LAA22466; Mon, 29 Jun 1998 11:46:48 +0200 (SAT) From: John Hay Message-Id: <199806290946.LAA22466@zibbi.mikom.csir.co.za> Subject: Re: time problem? In-Reply-To: <9570.899103386@critter.freebsd.dk> from Poul-Henning Kamp at "Jun 29, 98 08:56:26 am" To: phk@critter.freebsd.dk (Poul-Henning Kamp) Date: Mon, 29 Jun 1998 11:46:48 +0200 (SAT) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >I have an 1 pps signal connected to DCD of my serial port. > > What source do you use for this ? A Garmin GPS-25XL. I feed the signal to 2 machines, one is this SMP machine and the other is a 66MHz 486DX2 also running FreeBSD-current, although normally not quite as current. :-) I don't see those .85 second jumps on it, so I doubt if it is the pps signal. > > >---------------- > >A: 899000790.981073054 #29011 D: 1.000004191 > >A: 899000792.826104370 #29012 D: 1.845031316 > >A: 899016109.813482765 #44329 D: 0.999940488 > >A: 899016111.668519664 #44330 D: 1.855036899 > >---------- > > > >So on sequence numbers 29012 and 44330, I have gained .85 seconds > >in the space of a second. For the 15000 seconds between 29012 and > >44329 there were no glitches or jumps. The machine has been running for more than 12 hours now with your patch and there has been 5 .85 second jumps in that period. > > Now, one source of this could be interrupts diasabled for too long, > since your timestamps depend on the sio interrupt that would be my > primary suspect. I'll see if it still happen if I change the program not to use the sio timestamping, but just use one of the normal userlevel calls like gettimeofday() or clock_gettime(). > > Is there anything that makes it happen more often ? Not that I know of yet. I have left the machine idle for long periods and also done a lot of make worlds, but it did not seem to make a difference. > > >Is this function, i8254_get_timecount(), used by something that can > >have a permanent effect on time? I thought it was only used by > >nanotime() and microtime() to get the offset from the previous tick, > >but that it wasn't used to calculate the next tick? > > It shouldn't be, apart from "tco_forward()" which bumps the timecounters > reference time once each hardclock() (kern/kern_clock.c) Ok, I missed tco_delta() being called from sync_other_counter() at the start of tco_forward(); John -- John Hay -- John.Hay@mikom.csir.co.za To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 06:52:18 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA18509 for freebsd-smp-outgoing; Mon, 29 Jun 1998 06:52:18 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from dworkin.amber.org (petrilli@dworkin.amber.org [209.31.146.74]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA18497 for ; Mon, 29 Jun 1998 06:52:14 -0700 (PDT) (envelope-from petrilli@dworkin.amber.org) Received: from localhost (petrilli@localhost) by dworkin.amber.org (8.9.0/8.9.0) with SMTP id JAA28668; Mon, 29 Jun 1998 09:52:05 -0400 (EDT) Date: Mon, 29 Jun 1998 09:52:05 -0400 (EDT) From: "Christopher G. Petrilli" To: Atipa cc: Niall Smart , freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sun, 28 Jun 1998, Atipa wrote: > The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > utilize 100MHz system bus. This is a red herring... the system bus was a big restriction back when the cache was running at bus speed---any increment made a huge difference, but with the current PII architecture, the bus speed has long since ceased to be a problem. How many devices do you know that can saturate a PCI bus constitently? Don't use this as a reasoning. > Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock > rates, and much faster DRAM access, you'll definitely notice the > difference. Pros are at the end of their lifecycle, and will be hard to > support. The PROs will be faster ata given clock speed, and with the cache architecture of the Pros, probably at 50% above that, given a normal model of execution (80-90% cache hit rate). Remember, that you can get PPros with 512K or 1Mb of L2 cache that is running 1:1 with the chip, rather than 2:1. The big problem in most systems in my view (having been responsible for system optimization of mainframes and large UNIX boxes) is never the CPU---unless you're doing scientific applications, then it's usually memory---it's almost always I/O bandwidth. A faster disk, more memory for caching, more SCSI busses, etc, will make a BIG impact on your overal system throughput---which is a much better measure than "boboMIPS" :-) Just cuz you're CPU can spin faster, doesn't mean it doesn't do anything useful during those cycles. Chris -- | Christopher Petrilli | petrilli@amber.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 07:00:24 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA19715 for freebsd-smp-outgoing; Mon, 29 Jun 1998 07:00:24 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-03.waterford.indigo.ie [194.125.139.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA19673 for ; Mon, 29 Jun 1998 07:00:09 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id OAA00515; Mon, 29 Jun 1998 14:34:23 +0100 (IST) (envelope-from rotel@ginseng.indigo.ie) From: Niall Smart Message-Id: <199806291334.OAA00515@indigo.ie> Date: Mon, 29 Jun 1998 14:34:23 +0000 In-Reply-To: Atipa "Re: PPro vs PII" (Jun 28, 8:38pm) Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: Atipa , Niall Smart Subject: Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Jun 28, 8:38pm, Atipa wrote: } Subject: Re: PPro vs PII > > The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > utilize 100MHz system bus. I was thinking of trying 300Mhz P2s at 100Mhz system bus using the hack described at www.sysdoc.pair.com. > Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock > rates, and much faster DRAM access, you'll definitely notice the > difference. Pros are at the end of their lifecycle, and will be hard to > support. But the P2 has a much slower cache clock speed, what is a DIB? And how is the DRAM access faster if both the P2 and PPro use a 66Mhz system bus? Thanks, Niall -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org Annoy your enemies and astonish your friends: echo "#define if(x) if (!(x))" >> /usr/include/stdio.h To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 09:14:41 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA12901 for freebsd-smp-outgoing; Mon, 29 Jun 1998 09:14:41 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from bsd.synx.com (rt.synx.com [194.167.81.239]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id JAA12890 for ; Mon, 29 Jun 1998 09:14:32 -0700 (PDT) (envelope-from <@rn.synx.com:root@synx.com>) Received: from s3.synx.com (s3 [192.1.1.247]) by bsd.synx.com (8.6.12/8.6.12) with SMTP id RAA09419; Mon, 29 Jun 1998 17:14:24 +0100 Received: from rn by s3.synx.com id aa03736; 29 Jun 98 18:02 BST Date: Mon, 29 Jun 1998 18:14:13 +0200 (CEST) From: Remy NONNENMACHER Reply-To: remy@synx.com Subject: Re: PPro vs PII To: petrilli@dworkin.amber.org cc: freebsd@atipa.com, rotel@indigo.ie, freebsd-smp@FreeBSD.ORG In-Reply-To: MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Message-ID: <9806291802.aa03736@s3.synx.com> Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 29 Jun, Christopher G. Petrilli wrote: > On Sun, 28 Jun 1998, Atipa wrote: > >> The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can >> utilize 100MHz system bus. > > This is a red herring... the system bus was a big restriction back when > the cache was running at bus speed---any increment made a huge > difference, but with the current PII architecture, the bus speed has > long since ceased to be a problem. How many devices do you know that > can saturate a PCI bus constitently? Don't use this as a reasoning. > .... > model of execution (80-90% cache hit rate). Remember, that you can get > PPros with 512K or 1Mb of L2 cache that is running 1:1 with the chip, > rather than 2:1. > That's what say Tom's hardware guide: 100Mhz bus give a 3 to 10% increase. (Now, he uses Win...... benchmarks and it's difficult to figure out what is really tested.) Let's stop the pro/cons war about P2/Pro. I propose that those interested grab the rc5des client on www.distributed.net and post the result of -benchmark, along with processor type and speed. This will close the debate quickly. RN. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 11:30:50 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA08569 for freebsd-smp-outgoing; Mon, 29 Jun 1998 11:30:50 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from smtp04.primenet.com (daemon@smtp04.primenet.com [206.165.6.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA08473 for ; Mon, 29 Jun 1998 11:30:33 -0700 (PDT) (envelope-from tlambert@usr01.primenet.com) Received: (from daemon@localhost) by smtp04.primenet.com (8.8.8/8.8.8) id LAA16760; Mon, 29 Jun 1998 11:30:12 -0700 (MST) Received: from usr01.primenet.com(206.165.6.201) via SMTP by smtp04.primenet.com, id smtpd016670; Mon Jun 29 11:30:02 1998 Received: (from tlambert@localhost) by usr01.primenet.com (8.8.5/8.8.5) id LAA20596; Mon, 29 Jun 1998 11:29:57 -0700 (MST) From: Terry Lambert Message-Id: <199806291829.LAA20596@usr01.primenet.com> Subject: Re: PPro vs PII To: petrilli@dworkin.amber.org (Christopher G. Petrilli) Date: Mon, 29 Jun 1998 18:29:57 +0000 (GMT) Cc: freebsd@atipa.com, rotel@indigo.ie, freebsd-smp@FreeBSD.ORG In-Reply-To: from "Christopher G. Petrilli" at Jun 29, 98 09:52:05 am X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > This is a red herring... [ ... ] > How many devices do you know that can saturate a PCI bus constitently? Only the good ones. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 12:08:31 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA17473 for freebsd-smp-outgoing; Mon, 29 Jun 1998 12:08:31 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: (from sos@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA17283; Mon, 29 Jun 1998 12:07:48 -0700 (PDT) (envelope-from sos) Message-Id: <199806291907.MAA17283@hub.freebsd.org> Subject: Re: PPro vs PII In-Reply-To: <9806291802.aa03736@s3.synx.com> from Remy NONNENMACHER at "Jun 29, 98 06:14:13 pm" To: remy@synx.com Date: Mon, 29 Jun 1998 12:07:44 -0700 (PDT) Cc: petrilli@dworkin.amber.org, freebsd@atipa.com, rotel@indigo.ie, freebsd-smp@FreeBSD.ORG From: sos@FreeBSD.ORG Reply-to: sos@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In reply to Remy NONNENMACHER who wrote: > > That's what say Tom's hardware guide: 100Mhz bus give a 3 to 10% > increase. (Now, he uses Win...... benchmarks and it's difficult to > figure out what is really tested.) Sounds realistic, and close to my "real world" observations. > Let's stop the pro/cons war about P2/Pro. I propose that those > interested grab the rc5des client on www.distributed.net and post the > result of -benchmark, along with processor type and speed. This will > close the debate quickly. That will show us nothing, it will say nothing about how the machine will perform in real-life situations, its about as meningfull as the old dhrystone "benchmark"... -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Soren Schmidt (sos@FreeBSD.org) FreeBSD Core Team So much code to hack -- so little time. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 12:38:02 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA22503 for freebsd-smp-outgoing; Mon, 29 Jun 1998 12:38:02 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id MAA22471 for ; Mon, 29 Jun 1998 12:37:53 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 7896 invoked by uid 1017); 29 Jun 1998 18:34:50 -0000 Date: Mon, 29 Jun 1998 12:34:50 -0600 (MDT) From: Atipa To: Niall Smart cc: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: <199806291334.OAA00515@indigo.ie> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mon, 29 Jun 1998, Niall Smart wrote: > On Jun 28, 8:38pm, Atipa wrote: > } Subject: Re: PPro vs PII > > > > The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > > utilize 100MHz system bus. > > I was thinking of trying 300Mhz P2s at 100Mhz system bus using the hack > described at www.sysdoc.pair.com. Good plan, but don't let your vendor know. :) I maintain an anti-overclocking position at all times, because I think the headaches (for yourself and your vendors) exceed the performance gain. I only overclock (temporarily) game machines and VGA cards in said machines. No way would I even consider oc'ing a server... > > Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock > > rates, and much faster DRAM access, you'll definitely notice the > > difference. Pros are at the end of their lifecycle, and will be hard to > > support. > > But the P2 has a much slower cache clock speed, what is a DIB? Not really. On a clock for clock ratio, yes. But since the P2 clock rates are so much higher, the cache difference becomes less noticable. A P2-350 caches at 175, compared to 200 for a Pro. The DIB on the P2 allows higher hit rates and higher bandwidth between the L1, L2, and CPU. > And how is the DRAM access faster if both the P2 and PPro use a 66Mhz > system bus? The P2's can use a 440BX chipset, which supports 100MHz Sync DRAM (7ns), while the best production Pro chipset is the 440NX (Natoma), which supports only 66MHz EDO (async) DRAM at 60ns. EDO is 200MB/sec, while 100MHz SDRAM is over 500MB/sec. Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 12:45:24 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA23940 for freebsd-smp-outgoing; Mon, 29 Jun 1998 12:45:24 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id MAA23913 for ; Mon, 29 Jun 1998 12:45:12 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 7910 invoked by uid 1017); 29 Jun 1998 18:42:11 -0000 Date: Mon, 29 Jun 1998 12:42:11 -0600 (MDT) From: Atipa To: "Christopher G. Petrilli" cc: Niall Smart , freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > > utilize 100MHz system bus. > > This is a red herring... the system bus was a big restriction back when > the cache was running at bus speed---any increment made a huge > difference, but with the current PII architecture, the bus speed has > long since ceased to be a problem. How many devices do you know that > can saturate a PCI bus constitently? Don't use this as a reasoning. The bus can easily get saturated, and this is a _big_ difference, considering the DRAM, L2, PCI, AGP, etc., all are directly related to system bus speed. You yourself later admit that I/O usually binds system speed, and the system bus is a large portion. You can saturate a PCI bus with any one of the following: o (1) gigabit ethernet o (1) Matrox Meteor o (1) Adaptec 7890 Ultra-2 SCSI With any _one_ of those devices, excluding IDE, normal 100MBit ethernet, anf video (which is _very_heavy_) > > Since the P2 has DIB (dual independent bus) for the L2 cache, higher clock > > rates, and much faster DRAM access, you'll definitely notice the > > difference. Pros are at the end of their lifecycle, and will be hard to > > support. > > The PROs will be faster ata given clock speed, and with the cache > architecture of the Pros, probably at 50% above that, given a normal > model of execution (80-90% cache hit rate). Remember, that you can get > PPros with 512K or 1Mb of L2 cache that is running 1:1 with the chip, > rather than 2:1. But we are not talking philosophy here; who cares about clock-for-clock? This is the real world, where net performance is more important that theoretical (clock-for-clock) benchmarks. > The big problem in most systems in my view (having been responsible for > system optimization of mainframes and large UNIX boxes) is never the > CPU---unless you're doing scientific applications, then it's usually > memory---it's almost always I/O bandwidth. A faster disk, more memory > for caching, more SCSI busses, etc, will make a BIG impact on your > overal system throughput-- which a 100MHz bus will improve by 50%... > -which is a much better measure than > "boboMIPS" :-) Just cuz you're CPU can spin faster, doesn't mean it > doesn't do anything useful during those cycles. Correct. I would invest in a nice CCD setup w/ the Adpatec Ultra-2. But this is the -SMP list, so these are people that DO NEED the extra CPU. Probably breaking keys... ;) Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 13:08:11 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA27905 for freebsd-smp-outgoing; Mon, 29 Jun 1998 13:08:11 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-03.waterford.indigo.ie [194.125.139.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA27839 for ; Mon, 29 Jun 1998 13:07:46 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id VAA02803; Mon, 29 Jun 1998 21:03:24 +0100 (IST) (envelope-from rotel@indigo.ie) From: Niall Smart Message-Id: <199806292003.VAA02803@indigo.ie> Date: Mon, 29 Jun 1998 21:03:22 +0000 In-Reply-To: Atipa "Re: PPro vs PII" (Jun 29, 12:34pm) Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: Atipa , Niall Smart Subject: Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Jun 29, 12:34pm, Atipa wrote: } Subject: Re: PPro vs PII > > > And how is the DRAM access faster if both the P2 and PPro use a 66Mhz > > system bus? > > The P2's can use a 440BX chipset, which supports 100MHz Sync DRAM (7ns), > while the best production Pro chipset is the 440NX (Natoma), which > supports only 66MHz EDO (async) DRAM at 60ns. EDO is 200MB/sec, while > 100MHz SDRAM is over 500MB/sec. Yes, but benchmarks at tomshardware.com have already established that the 100Mhz memory bus offers little improvement over the 66Mhz bus on a PII. Niall -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 13:15:11 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA29268 for freebsd-smp-outgoing; Mon, 29 Jun 1998 13:15:11 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail1.its.rpi.edu (root@mail1.its.rpi.edu [128.113.100.7]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA29197 for ; Mon, 29 Jun 1998 13:14:42 -0700 (PDT) (envelope-from drosih@rpi.edu) Received: from [128.113.24.47] (gilead.acs.rpi.edu [128.113.24.47]) by mail1.its.rpi.edu (8.8.8/8.8.6) with ESMTP id QAA30170 for ; Mon, 29 Jun 1998 16:14:24 -0400 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Sender: drosih@pop1.rpi.edu Message-Id: In-Reply-To: <9806291802.aa03736@s3.synx.com> References: Date: Mon, 29 Jun 1998 16:18:12 -0400 To: freebsd-smp@FreeBSD.ORG From: Garance A Drosihn Subject: Re: PPro vs PII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org At 6:14 PM +0200 6/29/98, Remy NONNENMACHER wrote: > Let's stop the pro/cons war about P2/Pro. I propose that those > interested grab the rc5des client on www.distributed.net and post > the result of -benchmark, along with processor type and speed. > This will close the debate quickly. It will only close the debate on which machine is faster at running rc5des... It also doesn't address price/performance considerations. I went with a dual PPro system simply because I stumbled upon a great deal. On the other hand, I got that great deal because PPro's are getting somewhat scarce, so I can understand someone wanting to buy into the newer chipsets. --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer or drosih@rpi.edu Rensselaer Polytechnic Institute To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 15:47:28 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA26375 for freebsd-smp-outgoing; Mon, 29 Jun 1998 15:47:28 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from compound.east.sun.com ([208.141.230.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA26350 for ; Mon, 29 Jun 1998 15:47:05 -0700 (PDT) (envelope-from alk@compound.east.sun.com) Received: (from alk@localhost) by compound.east.sun.com (8.8.8/8.7.3) id RAA04842; Mon, 29 Jun 1998 17:47:58 -0500 (CDT) From: Tony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Mon, 29 Jun 1998 17:47:57 -0500 (CDT) X-Face: O9M"E%K;(f-Go/XDxL+pCxI5*gr[=FN@Y`cl1.Tn Reply-To: alk@pobox.com To: rotel@indigo.ie Cc: freebsd@atipa.com, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <199806292003.VAA02803@indigo.ie> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <13720.5928.221514.597576@compound.east> Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Quoth Niall Smart on Mon, 29 June: : : Yes, but benchmarks at tomshardware.com have already established : that the 100Mhz memory bus offers little improvement over the 66Mhz : bus on a PII. Kinda sad, to be squabbling about the tiny differences between PII/PPro, when G3, Alpha, and UltraSparc systems are whipping the '86s by much more substantial margins. What's SpecInt95/SpecFP95 on a 450MHz G3, anyhow? I know USparc IIi is more than twice as fast as a PII on my FP kernels. Can you get Alphas from Samsung yet? That should be a commodity-pricing watershed event, when it occurs. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 17:07:35 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA09877 for freebsd-smp-outgoing; Mon, 29 Jun 1998 17:07:35 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from homer.supersex.com (homer.supersex.com [209.5.1.4]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA09818 for ; Mon, 29 Jun 1998 17:07:02 -0700 (PDT) (envelope-from leo@homer.supersex.com) Received: (from leo@localhost) by homer.supersex.com (8.8.8/8.8.5) id UAA19114; Mon, 29 Jun 1998 20:07:54 -0400 (EDT) Message-ID: <19980629200753.14325@supersex.com> Date: Mon, 29 Jun 1998 20:07:53 -0400 From: Leo Papandreou To: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <9806291802.aa03736@s3.synx.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <9806291802.aa03736@s3.synx.com>; from Remy NONNENMACHER on Mon, Jun 29, 1998 at 06:14:13PM +0200 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mon, Jun 29, 1998 at 06:14:13PM +0200, Remy NONNENMACHER wrote: > On 29 Jun, Christopher G. Petrilli wrote: > > On Sun, 28 Jun 1998, Atipa wrote: > > > >> The P2 will smoke it. Better yet, go up to 350 or 400MHz, then you can > >> utilize 100MHz system bus. > > > > This is a red herring... the system bus was a big restriction back when > > the cache was running at bus speed---any increment made a huge > > difference, but with the current PII architecture, the bus speed has > > long since ceased to be a problem. How many devices do you know that > > can saturate a PCI bus constitently? Don't use this as a reasoning. > > .... > > model of execution (80-90% cache hit rate). Remember, that you can get > > PPros with 512K or 1Mb of L2 cache that is running 1:1 with the chip, > > rather than 2:1. > > > > That's what say Tom's hardware guide: 100Mhz bus give a 3 to 10% > increase. (Now, he uses Win...... benchmarks and it's difficult to > figure out what is really tested.) True. I dont know how many processes are running in the background of a typical windows session but I imagine one person is running one application which is spending nearly all of its time in the CPU cache. When you've spawned upwards of 200 httpd daemons and whatever else not in ram, I bet the "3 to 10%" improvement improves. This is precisely the case where I admin. There is also HEAVY file io spread out across literally hundreds of thousands of different image files which need to be read, stuffed into a socket and often processed to generate thumb- nails on the fly. Whatever was in the CPU cache isnt going to be there next time you blink. All of which to say, Tom's experience with Quake is not necessarily indicative of what a 100Mz bus can do. Fast wide buses are just the next best thing since goldfish. Given that newer PIIs/mobos will soon be running their cache at full speed, just like the Pro, I'd say the Pro is not worth looking at if you can swallow the difference in price for a PII. > > Let's stop the pro/cons war about P2/Pro. I propose that those > interested grab the rc5des client on www.distributed.net and post the > result of -benchmark, along with processor type and speed. This will > close the debate quickly. > Heh. A little anecdotal evidence: we upgraded from a P150 to a PII400/BX combination and believe you me, there's a difference (I just like bragging about how far I can pee :-) I can hardly wait for -CURRENT's support of 2 or more PIIs to become the next -STABLE. > RN. > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-smp" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 17:17:57 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA11185 for freebsd-smp-outgoing; Mon, 29 Jun 1998 17:17:57 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from homer.supersex.com (homer.supersex.com [209.5.1.4]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA11113 for ; Mon, 29 Jun 1998 17:17:17 -0700 (PDT) (envelope-from leo@homer.supersex.com) Received: (from leo@localhost) by homer.supersex.com (8.8.8/8.8.5) id UAA19182; Mon, 29 Jun 1998 20:17:58 -0400 (EDT) Message-ID: <19980629201758.52413@supersex.com> Date: Mon, 29 Jun 1998 20:17:58 -0400 From: Leo Papandreou To: rotel@indigo.iey Cc: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <199806292003.VAA02803@indigo.ie> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89i In-Reply-To: <199806292003.VAA02803@indigo.ie>; from Niall Smart on Mon, Jun 29, 1998 at 09:03:22PM +0000 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Mon, Jun 29, 1998 at 09:03:22PM +0000, Niall Smart wrote: > On Jun 29, 12:34pm, Atipa wrote: > } Subject: Re: PPro vs PII > > > > > And how is the DRAM access faster if both the P2 and PPro use a 66Mhz > > > system bus? > > > > The P2's can use a 440BX chipset, which supports 100MHz Sync DRAM (7ns), > > while the best production Pro chipset is the 440NX (Natoma), which > > supports only 66MHz EDO (async) DRAM at 60ns. EDO is 200MB/sec, while > > 100MHz SDRAM is over 500MB/sec. AND, PIIs will soon run much larger caches at full speed. Contest over. > > Yes, but benchmarks at tomshardware.com have already established > that the 100Mhz memory bus offers little improvement over the 66Mhz > bus on a PII. > Fact: the benchmarks at tomshardware.com have established that the 100Mz memory bus offers little improvement over the 66Mz in running the benchmarks at tomshardware.com. Tom benchmarks Windows stuff. Pushing the mouse around some empty real estate on a Windows95 screen causes CPU usage to jump past 60%. > Niall > > -- > Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk > FreeBSD: Turning PC's into Workstations: www.freebsd.org > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-smp" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 19:13:27 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id TAA02588 for freebsd-smp-outgoing; Mon, 29 Jun 1998 19:13:27 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-03.waterford.indigo.ie [194.125.139.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id TAA02551 for ; Mon, 29 Jun 1998 19:13:14 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id DAA03893; Tue, 30 Jun 1998 03:09:09 +0100 (IST) (envelope-from rotel@indigo.ie) From: Niall Smart Message-Id: <199806300209.DAA03893@indigo.ie> Date: Tue, 30 Jun 1998 03:09:09 +0000 In-Reply-To: Leo Papandreou "Re: PPro vs PII" (Jun 29, 8:17pm) Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: Leo Papandreou , rotel@indigo.iey Subject: Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Jun 29, 8:17pm, Leo Papandreou wrote: } Subject: Re: PPro vs PII > On Mon, Jun 29, 1998 at 09:03:22PM +0000, Niall Smart wrote: > > On Jun 29, 12:34pm, Atipa wrote: > > } Subject: Re: PPro vs PII > > > > > > > And how is the DRAM access faster if both the P2 and PPro use a 66Mhz > > > > system bus? > > > > > > The P2's can use a 440BX chipset, which supports 100MHz Sync DRAM (7ns), > > > while the best production Pro chipset is the 440NX (Natoma), which > > > supports only 66MHz EDO (async) DRAM at 60ns. EDO is 200MB/sec, while > > > 100MHz SDRAM is over 500MB/sec. > > AND, PIIs will soon run much larger caches at full speed. Contest over. At exorbitant prices though, a dual PPro board *with* 2 180/256 chips, uw scsi, intel 10/100 net and sb sound can bve had for ~$320. Those chips aren't exactly fast, but you can move up to 200/512 if you like and clock them at 233. Of course the xeon will whup its ass, but it doesn't come close on price/performance. Anyway the original question I asked was how such a system would compare to a dual PII 300. > > Yes, but benchmarks at tomshardware.com have already established > > that the 100Mhz memory bus offers little improvement over the 66Mhz [snip] > Tom benchmarks Windows stuff. Pushing the mouse around some empty > real estate on a Windows95 screen causes CPU usage to jump past 60%. Bah, I think the guy knows what he is doing. If x is faster than y under NT it should hold for FreeBSD. Niall -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 19:14:26 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id TAA02756 for freebsd-smp-outgoing; Mon, 29 Jun 1998 19:14:26 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-03.waterford.indigo.ie [194.125.139.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id TAA02666 for ; Mon, 29 Jun 1998 19:14:00 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id DAA03914; Tue, 30 Jun 1998 03:10:00 +0100 (IST) (envelope-from rotel@indigo.ie) From: Niall Smart Message-Id: <199806300210.DAA03914@indigo.ie> Date: Tue, 30 Jun 1998 03:10:00 +0000 In-Reply-To: Garance A Drosihn "Re: PPro vs PII" (Jun 29, 4:18pm) Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: Garance A Drosihn , freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Jun 29, 4:18pm, Garance A Drosihn wrote: } Subject: Re: PPro vs PII > > It also doesn't address price/performance considerations. I went > with a dual PPro system simply because I stumbled upon a great > deal. On the other hand, I got that great deal because PPro's any left? -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 19:31:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id TAA06635 for freebsd-smp-outgoing; Mon, 29 Jun 1998 19:31:09 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.xmission.com (mail.xmission.com [198.60.22.22]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id TAA06620 for ; Mon, 29 Jun 1998 19:31:00 -0700 (PDT) (envelope-from seth@pengar.com) Received: from ([192.168.0.2]) [166.70.2.84] by mail.xmission.com with esmtp (Exim 1.82 #2) id 0yqqC7-00002j-00; Mon, 29 Jun 1998 20:30:56 -0600 X-Sender: seth-pc@hobbiton.shire.net Message-Id: In-Reply-To: <9806291802.aa03736@s3.synx.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Mon, 29 Jun 1998 20:33:57 +0100 To: freebsd-smp@FreeBSD.ORG From: Seth Leigh Subject: Re: PPro vs PII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Intel Providence (PR440FX) mobo, one 64 MB DIMM Two Pentium Pro 180/256 overclocked (solid as rocks) to 200 MHz FreeBSD 3.0-031198-SNAP Running two RC5 clients, each in its own xterm, each process averages around 526 kKeys/sec sustained, for a combined key rate of around 1056 kKeys/sec. If I don't run X Windows (or anything else for that matter) and run them in two virtual terminals, I have seen both processes average 537 kKeys/sec. This is not bad at all (my single-cpu Cyrix P166+ gets 309 kKeys/sec), BUT... I believe fully that a dual PII-300 system (as the original poster was questioning about) would handily beat this key rate. Probably by around 50% ;-). The benchmarks show that a single-cpu PII-400 system will just barely (by up to 50 kKeys/sec) beat the combined output of my dual PPro 200 system. You have to remember that the RC5 client uses a code core and data chunk so small that they easily fit in the cpu's internal cache, so RC5 client speeds are NOT a good general benchmark. Seth >Let's stop the pro/cons war about P2/Pro. I propose that those >interested grab the rc5des client on www.distributed.net and post the >result of -benchmark, along with processor type and speed. This will >close the debate quickly. > >RN. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 20:50:43 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA20021 for freebsd-smp-outgoing; Mon, 29 Jun 1998 20:50:43 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id UAA19937 for ; Mon, 29 Jun 1998 20:50:21 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 9570 invoked by uid 1017); 30 Jun 1998 02:47:20 -0000 Date: Mon, 29 Jun 1998 20:47:20 -0600 (MDT) From: Atipa To: Leo Papandreou cc: freebsd-smp@FreeBSD.ORG, sos@FreeBSD.ORG Subject: Intel announces Xeon!!! Was Re: PPro vs PII In-Reply-To: <19980629200753.14325@supersex.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org This thread is now obselete, since the new Xeon has the best features of both chips. Features Comment ------------------- -------------------- 400Mhz CPU Fast as fastest P2 400MHz cache 1:1 caching (like PPro) DIB Dual Independent Bus (from P2; double bandwidth) 1MB L2 cache From PPro. 36-bit memory Supports over 64GB RAM cached (best yet) 8-way SMP capable Better than both (w/o weird chipsets) "Cluster Support" Can cluster 4 servers? Ask Intel on that... Thermal Sensor Integrated into S.E.C., microcode ECC L2 cache Only on faster P2's right now FRC checking Redundant checks for SMP tasks (master/slave) System Management Bus S.E.C. firmware for remote diagnostics / maint... Atipa will be carrying them very shortly. Check our site for pricing and availability. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 20:52:52 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA20281 for freebsd-smp-outgoing; Mon, 29 Jun 1998 20:52:52 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id UAA20252 for ; Mon, 29 Jun 1998 20:52:37 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 9583 invoked by uid 1017); 30 Jun 1998 02:49:34 -0000 Date: Mon, 29 Jun 1998 20:49:34 -0600 (MDT) From: Atipa To: Leo Papandreou cc: freebsd-smp@FreeBSD.ORG, sos@FreeBSD.ORG Subject: Re: Intel announces Xeon!!! Was Re: PPro vs PII In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Forgot the URL for the Press Release: http://www.intel.com/PentiumII/Xeon/intro.htm?iid=mail+cn2& > This thread is now obselete, since the new Xeon has the best features of > both chips. > > Features Comment > ------------------- -------------------- > 400Mhz CPU Fast as fastest P2 > 400MHz cache 1:1 caching (like PPro) > DIB Dual Independent Bus (from P2; double bandwidth) > 1MB L2 cache From PPro. > 36-bit memory Supports over 64GB RAM cached (best yet) > 8-way SMP capable Better than both (w/o weird chipsets) > "Cluster Support" Can cluster 4 servers? Ask Intel on that... > Thermal Sensor Integrated into S.E.C., microcode > ECC L2 cache Only on faster P2's right now > FRC checking Redundant checks for SMP tasks (master/slave) > System Management Bus S.E.C. firmware for remote diagnostics / maint... > > Atipa will be carrying them very shortly. Check our site for pricing and > availability. > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 21:15:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA23437 for freebsd-smp-outgoing; Mon, 29 Jun 1998 21:15:09 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id VAA23429 for ; Mon, 29 Jun 1998 21:15:03 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 9715 invoked by uid 1017); 30 Jun 1998 03:12:03 -0000 Date: Mon, 29 Jun 1998 21:12:03 -0600 (MDT) From: Atipa To: Leo Papandreou cc: freebsd-smp@FreeBSD.ORG, sos@FreeBSD.ORG Subject: Motherboard for Xeon Processor In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org The new chip needs a new chipset and motherboard (Slot 2 SEC). Here are the details of the first board to support the new architecture: http://developer.intel.com/design/wrkstn/MS440gx/index.htm It is a fairly nice 2-way SMP board w/ 5 PCI slots + 1 AGP, supporting up to 2GB DRAM. Their new bus handles up to 800MB/sec between CPUS and caches. Also has integrated Crystal Audio, and EtherExpress Pro 100Mbit ethernet. Prices and availability to follow. Kevin On Mon, 29 Jun 1998, Atipa wrote: > > Forgot the URL for the Press Release: > http://www.intel.com/PentiumII/Xeon/intro.htm?iid=mail+cn2& > > > This thread is now obselete, since the new Xeon has the best features of > > both chips. > > > > Features Comment > > ------------------- -------------------- > > 400Mhz CPU Fast as fastest P2 > > 400MHz cache 1:1 caching (like PPro) > > DIB Dual Independent Bus (from P2; double bandwidth) > > 1MB L2 cache From PPro. > > 36-bit memory Supports over 64GB RAM cached (best yet) > > 8-way SMP capable Better than both (w/o weird chipsets) > > "Cluster Support" Can cluster 4 servers? Ask Intel on that... > > Thermal Sensor Integrated into S.E.C., microcode > > ECC L2 cache Only on faster P2's right now > > FRC checking Redundant checks for SMP tasks (master/slave) > > System Management Bus S.E.C. firmware for remote diagnostics / maint... > > > > Atipa will be carrying them very shortly. Check our site for pricing and > > availability. > > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 21:28:00 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA25832 for freebsd-smp-outgoing; Mon, 29 Jun 1998 21:28:00 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id VAA25794 for ; Mon, 29 Jun 1998 21:27:47 -0700 (PDT) (envelope-from freebsd@atipa.com) Received: (qmail 9741 invoked by uid 1017); 30 Jun 1998 03:24:47 -0000 Date: Mon, 29 Jun 1998 21:24:47 -0600 (MDT) From: Atipa To: Tony Kimball cc: rotel@indigo.ie, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: <13720.5928.221514.597576@compound.east> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > : Yes, but benchmarks at tomshardware.com have already established > : that the 100Mhz memory bus offers little improvement over the 66Mhz > : bus on a PII. > > Kinda sad, to be squabbling about the tiny differences between > PII/PPro, when G3, Alpha, and UltraSparc systems are whipping the > '86s by much more substantial margins. What's SpecInt95/SpecFP95 on a > 450MHz G3, anyhow? The SPECfp95 for their new 400MHz chip w/ 1MB L2 (400MHz) is 13.7, compared to 23.5 for a Sun Ultra60 Model 360. You can use up to 8 Xeons simultaneously. > I know USparc IIi is more than twice as fast as a > PII on my FP kernels. Can you get Alphas from Samsung yet? That > should be a commodity-pricing watershed event, when it occurs. That would be nice. I am worried about the future of Alphas at this point... Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 21:59:12 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id VAA29459 for freebsd-smp-outgoing; Mon, 29 Jun 1998 21:59:12 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from compound.east.sun.com ([208.141.230.135]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id VAA29440; Mon, 29 Jun 1998 21:59:05 -0700 (PDT) (envelope-from alk@compound.east.sun.com) Received: (from alk@localhost) by compound.east.sun.com (8.8.8/8.7.3) id AAA06867; Tue, 30 Jun 1998 00:00:07 -0500 (CDT) From: Tony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 30 Jun 1998 00:00:03 -0500 (CDT) X-Face: O9M"E%K;(f-Go/XDxL+pCxI5*gr[=FN@Y`cl1.Tn Reply-To: alk@pobox.com To: freebsd@atipa.com Cc: chat@FreeBSD.ORG Subject: Re: PPro vs PII References: <13720.5928.221514.597576@compound.east> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <13720.27283.95382.801320@compound.east> Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'll move this to chat. Quoth Atipa on Mon, 29 June: : : You can use up to 8 Xeons simultaneously. You can have a box off-the-shelf with 64 USparc too. Heck, look at ASCI Red -- it proves by demonstration that you can buy a box to use up to 9152 PPros simultaneously -- if you don't mind some funky address space structure:) : > ...Samsung... : : That would be nice. I am worried about the future of Alphas at this : point... No fear! Merced got delayed *again*, (and probably will a couple more times before I retire to Costa Rica) and Alpha is the *only* commodity 64-bit platform (outside the embedded world) -- since I can't count USparc as commodity, even if it is at commodity pricing levels, as long as it is single-source. Merced might not dent the market appreciably until 2003! By which time Alpha and USparc will have dominated the large database world very nicely. If Compaq pushes Alpha in the Fortune 1000 server space like they say they will, and Samsung is producing Alphas at competetive-market pricing, and every EDA engineer on the planet would kill his boss in order to swap his Klamath/NT desktop box for an Alpha/Linux box (and a few disgruntled, newly promoted ones may start doing so in fact RSN -- if only metaphorically), Alpha is sitting prettier than any other ISA going. Remember -- x86 can only lose share at this point, not gain. Heck, every time SGI makes another management blunder, and that's about every 60 seconds lately, a few hundred more Alphas are, in effect, sold. Intel doesn't even have to lose share for Alpha to grow (although they will) -- there is a lot of consolidation going on in the Workstation (and by implication, the ISA) space. HP really should start feeling it's age before long. If they didn't have such unrealistic market cap, hung over from the mainframe days, they would have been bought by Dell or some such nouveau riche chop shop long ago. I'd feel pretty good buying Alpha hardware right about now if only there was an upgrade roadmap. (That's always been the illusory carrot motivating hobbyist hardware choices -- historically favoring intel -- hasn't it? Modular upgrades, carrying over hardware from generation to generation.) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 22:42:29 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA05101 for freebsd-smp-outgoing; Mon, 29 Jun 1998 22:42:29 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from cheops.anu.edu.au (bryan@cheops.anu.edu.au [150.203.76.24]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA05095 for ; Mon, 29 Jun 1998 22:42:26 -0700 (PDT) (envelope-from bryan@coombs.anu.edu.au) Message-Id: <199806300542.WAA05095@hub.freebsd.org> Received: by cheops.anu.edu.au (1.37.109.16/16.2) id AA177425325; Tue, 30 Jun 1998 15:42:05 +1000 From: bryan collins Subject: Re: PPro vs PII In-Reply-To: from Seth Leigh at "Jun 29, 98 08:33:57 pm" To: seth@pengar.com (Seth Leigh) Date: Tue, 30 Jun 1998 15:42:05 +1000 (EST) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL38 (25)] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Seth Leigh said: > Intel Providence (PR440FX) mobo, one 64 MB DIMM > Two Pentium Pro 180/256 overclocked (solid as rocks) to 200 MHz > FreeBSD 3.0-031198-SNAP > Running two RC5 clients, each in its own xterm, each process averages > around 526 kKeys/sec sustained, for a combined key rate of around 1056 > kKeys/sec. If I don't run X Windows (or anything else for that matter) and > run them in two virtual terminals, I have seen both processes average 537 > kKeys/sec. This is not bad at all (my single-cpu Cyrix P166+ gets 309 > kKeys/sec), BUT... > I have a dual PentiumII 350mhz on a SuperMicro P6DBS motherboard. I get 1.1million keys/sec with DES and 654k keys/sec with RC5 Thats two becnhmarks at once Bry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 23:14:12 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA08931 for freebsd-smp-outgoing; Mon, 29 Jun 1998 23:14:12 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from panzer.plutotech.com (ken@panzer.plutotech.com [206.168.67.125]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA08869; Mon, 29 Jun 1998 23:14:03 -0700 (PDT) (envelope-from ken@panzer.plutotech.com) Received: (from ken@localhost) by panzer.plutotech.com (8.8.8/8.8.5) id AAA26408; Tue, 30 Jun 1998 00:13:56 -0600 (MDT) From: "Kenneth D. Merry" Message-Id: <199806300613.AAA26408@panzer.plutotech.com> Subject: Re: Intel announces Xeon!!! Was Re: PPro vs PII In-Reply-To: from Atipa at "Jun 29, 98 08:47:20 pm" To: freebsd@atipa.com (Atipa) Date: Tue, 30 Jun 1998 00:13:56 -0600 (MDT) Cc: leo@talcom.net, freebsd-smp@FreeBSD.ORG, sos@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 Precedence: bulk X-Loop: FreeBSD.org Atipa wrote... > > This thread is now obselete, since the new Xeon has the best features of > both chips. > > Features Comment > ------------------- -------------------- > 400Mhz CPU Fast as fastest P2 > 400MHz cache 1:1 caching (like PPro) > DIB Dual Independent Bus (from P2; double bandwidth) > 1MB L2 cache From PPro. > 36-bit memory Supports over 64GB RAM cached (best yet) Although the 450NX chipset only support 8GB DRAM. The interesting thing is, the chipset supports 36-bit addressing. > 8-way SMP capable Better than both (w/o weird chipsets) Well, the Xeon itself only has "native" support for 4-way SMP. (despite what Intel's product brief says: http://developer.intel.com/design/pentiumii/xeon/prodbref/ Their data sheet for the Xeon: http://developer.intel.com/design/pentiumii/xeon/datashts/243770.htm indicates that it only has "native" support for 4-way SMP. To do otherwise would require changing the P6 processor bus, which they haven't done.) The 450NX chipset apparantly has provision for connecting a 3rd party "cluster bridge" to the system bus and the "MIOC" (memory and I/O controller). The thing they say is that the 3rd party cluster bridge would operate at "reduced frequencies". (see: http://developer.intel.com/design/pcisets/450nx/index.htm) Their "and beyond" statement on that page leads me to believe that you could probably have a cluster bridge that connects more than two P6 busses. For instance, you could have 8 groups of 450NX chipsets, each with 4 Xeon processors and 8GB of RAM, all connected by third-party processor bus bridges. Since the chips and the chipsets all support 36-bit addressing, you would be able to address all 64GB of memory from any one of the 32 processors. Sounds neat. > "Cluster Support" Can cluster 4 servers? Ask Intel on that... I think the "cluster support" refers to clusters of Xeon processors connected by processor bus bridge chips, like Corollary supposedly makes. I think Intel bought them. Axil (http://www.axil.com) already makes 8-way systems. But I think they've been pretty much shut down. See: http://www.eet.com/news/98/1014news/cuts.html Ken -- Kenneth Merry ken@plutotech.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Mon Jun 29 23:32:56 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA12447 for freebsd-smp-outgoing; Mon, 29 Jun 1998 23:32:56 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from tarsier.ca.sandia.gov (tarsier.ca.sandia.gov [146.246.246.124]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA12429 for ; Mon, 29 Jun 1998 23:32:53 -0700 (PDT) (envelope-from cc@tarsier.ca.sandia.gov) Received: from tarsier.ca.sandia.gov (localhost [127.0.0.1]) by tarsier.ca.sandia.gov (8.8.8/8.8.8) with ESMTP id XAA09584; Mon, 29 Jun 1998 23:30:21 -0700 (PDT) (envelope-from cc@tarsier.ca.sandia.gov) Message-Id: <199806300630.XAA09584@tarsier.ca.sandia.gov> X-Mailer: exmh version 2.0.2 2/24/98 To: Atipa cc: Tony Kimball , rotel@indigo.ie, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-reply-to: Your message of "Mon, 29 Jun 1998 21:24:47 MDT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 29 Jun 1998 23:30:21 -0700 From: "Chris Csanady" Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > >> : Yes, but benchmarks at tomshardware.com have already established >> : that the 100Mhz memory bus offers little improvement over the 66Mhz >> : bus on a PII. >> >> Kinda sad, to be squabbling about the tiny differences between >> PII/PPro, when G3, Alpha, and UltraSparc systems are whipping the >> '86s by much more substantial margins. What's SpecInt95/SpecFP95 on a >> 450MHz G3, anyhow? > >The SPECfp95 for their new 400MHz chip w/ 1MB L2 (400MHz) is 13.7, >compared to 23.5 for a Sun Ultra60 Model 360. Meanwhile, I think the new Alpha is safely keeping its lead. :) KP21264-3.5X Est. 43 SPECint95/64 SPECfp95 I must say, there is nothing like breakthrough after breakthrough from intel. Chris To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 00:33:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA22203 for freebsd-smp-outgoing; Tue, 30 Jun 1998 00:33:09 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA22156 for ; Tue, 30 Jun 1998 00:32:48 -0700 (PDT) (envelope-from jkh@time.cdrom.com) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.8/8.8.8) with ESMTP id AAA07164; Tue, 30 Jun 1998 00:32:14 -0700 (PDT) (envelope-from jkh@time.cdrom.com) To: Leo Papandreou cc: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-reply-to: Your message of "Mon, 29 Jun 1998 20:07:53 EDT." <19980629200753.14325@supersex.com> Date: Tue, 30 Jun 1998 00:32:13 -0700 Message-ID: <7160.899191933@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > Given that newer PIIs/mobos will soon be running their cache at full > speed, just like the Pro, I'd say the Pro is not worth looking at if > you can swallow the difference in price for a PII. Yes, but that's not going to help you now nor will you be able to preserve your investment in CPU technology if you go this route now, before the Xeon technology is released. I think that what people are trying to say right now is that a dual P6 system is REALLY CHEAP and a heck of a good way to get into the SMP waters for far less money than people used to spend on 286 machines. You can go out and get an Intel dual P6 Providence mobo with on-board SCSI WIDE and 100MBit ethernet support (AIC 7880 and Pro 100B respectively) for < $100 now (we've been getting them for $89) and the CPUs for $200-$300 each. If you can put together a full system including memory, CPUs and net/scsi ctlr/4GB disk for $1K right now in the PII camp then I'd be most interested to hear it. Since I think the answer is currently that "you can't", there's still considerable merit to the idea of buying 3 of these systems for $3K and deploying them far more favorably than you could a single system costing the same price. I also agree with the person who said that a P6/PII argument was senseless. It's really a price/performance issue and always has been for most people. - Jordan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 00:35:36 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA22522 for freebsd-smp-outgoing; Tue, 30 Jun 1998 00:35:36 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA22511 for ; Tue, 30 Jun 1998 00:35:23 -0700 (PDT) (envelope-from jkh@time.cdrom.com) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.8/8.8.8) with ESMTP id AAA07183; Tue, 30 Jun 1998 00:34:48 -0700 (PDT) (envelope-from jkh@time.cdrom.com) To: Leo Papandreou cc: rotel@indigo.iey, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-reply-to: Your message of "Mon, 29 Jun 1998 20:17:58 EDT." <19980629201758.52413@supersex.com> Date: Tue, 30 Jun 1998 00:34:47 -0700 Message-ID: <7178.899192087@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > AND, PIIs will soon run much larger caches at full speed. Contest over. You're seriously premature. The contest will be over when the fat lady sings and Intel actually SHIPS the parts they are already overdue on. If I wanted vaporware of either the hardware or software variety, I'd be buying microsoft products. :) - Jordan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 00:40:12 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA23143 for freebsd-smp-outgoing; Tue, 30 Jun 1998 00:40:12 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA23083 for ; Tue, 30 Jun 1998 00:39:31 -0700 (PDT) (envelope-from jkh@time.cdrom.com) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.8/8.8.8) with ESMTP id AAA07202; Tue, 30 Jun 1998 00:38:40 -0700 (PDT) (envelope-from jkh@time.cdrom.com) To: Atipa cc: Tony Kimball , rotel@indigo.ie, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-reply-to: Your message of "Mon, 29 Jun 1998 21:24:47 MDT." Date: Tue, 30 Jun 1998 00:38:40 -0700 Message-ID: <7197.899192320@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > You can use up to 8 Xeons simultaneously. I wish people would stop arguing the theoretical and start arguing the practical. Where can I buy an 8-way Xeon board? And for how much? And can I get 8 chips for it without any hassles? If the answer is "not yet" then Bzzzzzt, thanks for playing, and here's a copy of our home game. Come back when there's an actual, tangible product to talk about. :) P.S. You never did take up my challenge of proposing a better ports collection design after raising such a ruckus about it before. Do you also design, or only whine? :) - Jordan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 00:46:30 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA24352 for freebsd-smp-outgoing; Tue, 30 Jun 1998 00:46:30 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from proxy4.ba.best.com (root@proxy4.ba.best.com [206.184.139.15]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA24260 for ; Tue, 30 Jun 1998 00:45:58 -0700 (PDT) (envelope-from spadger@best.com) Received: from best.com (dynamic44.pm05.sf3d.best.com [209.24.235.44]) by proxy4.ba.best.com (8.8.8/8.8.BEST) with ESMTP id AAA17812; Tue, 30 Jun 1998 00:44:49 -0700 (PDT) Message-ID: <35989768.71FABE90@best.com> Date: Tue, 30 Jun 1998 00:44:40 -0700 From: Andy Sparrow Organization: Not likely. X-Mailer: Mozilla 4.05 [en] (X11; I; FreeBSD 3.0-CURRENT i386) MIME-Version: 1.0 To: bryan collins CC: Seth Leigh , freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <199806300542.WAA05095@hub.freebsd.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org bryan collins wrote: > Seth Leigh said: > > Intel Providence (PR440FX) mobo, one 64 MB DIMM > > Two Pentium Pro 180/256 overclocked (solid as rocks) to 200 MHz > > FreeBSD 3.0-031198-SNAP > > Running two RC5 clients, each in its own xterm, each process averages > > around 526 kKeys/sec sustained, for a combined key rate of around 1056 > > kKeys/sec. If I don't run X Windows (or anything else for that matter) and > > run them in two virtual terminals, I have seen both processes average 537 > > kKeys/sec. This is not bad at all (my single-cpu Cyrix P166+ gets 309 > > kKeys/sec), BUT... > > > > I have a dual PentiumII 350mhz on a SuperMicro P6DBS motherboard. > > I get 1.1million keys/sec with DES > and 654k keys/sec with RC5 > > Thats two becnhmarks at once Interesting. Micron W6-Li, (440FX "Natoma" chipset with integrated 2940UW & sound), fitted with dual P6 166/512s overclocked to 233Mhz, running -current from Jun 14th, 128Mb, 4Gb Seagate. Benchmarks at 1M key/sec with DES, 652k keys/sec with RC5, whilst running X, Netscrape, GIMP. Not too shabby for a sub $1000 machine assembled from brand-new components... (no, that figure doesn't include the #9 Imagine 128 with 8Mb or the 20" monitor - but it includes everything else.. ;-) But, as another writer pointed out, this only really proves how fast the machine runs the RC5 client, nothing else.... In *real life*, the system does anywhere from 617k keys/sec to 644 k keys/sec per CPU, depending upon load. It's up 24/7 until I do an "make installworld". Then it starts all over again :=) My personal experience with overclocking (4 machines, none running at the rated speed, all completely stable for months and months, and all significantly faster than stock) leads me to believe that there's nothing wrong with it - so long as you don't get silly, use thermal compound and efficient heatsinks/fans etc. After all, isn't ftp.cdrom.com, the biggest, busiest FTP server on the planet, an overclocked P6 @ 233Mhz? Even with 3200 concurrent connections, that machine doesn't seem to have much problem filling a T1 when I connect... Guess it all depends on how much you want to spend and what you want to do. Like someone else said - the PPro is hard to beat for price/performance ratio right now. Cheers, AS To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 01:00:05 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id BAA27022 for freebsd-smp-outgoing; Tue, 30 Jun 1998 01:00:05 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from proxy4.ba.best.com (root@proxy4.ba.best.com [206.184.139.15]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA26922 for ; Tue, 30 Jun 1998 00:59:29 -0700 (PDT) (envelope-from spadger@best.com) Received: from best.com (dynamic44.pm05.sf3d.best.com [209.24.235.44]) by proxy4.ba.best.com (8.8.8/8.8.BEST) with ESMTP id AAA22015; Tue, 30 Jun 1998 00:58:16 -0700 (PDT) Message-ID: <35989A91.BBF90786@best.com> Date: Tue, 30 Jun 1998 00:58:09 -0700 From: Andy Sparrow Organization: Not likely. X-Mailer: Mozilla 4.05 [en] (X11; I; FreeBSD 3.0-CURRENT i386) MIME-Version: 1.0 To: rotel@indigo.ie CC: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <199806300210.DAA03914@indigo.ie> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Niall Smart wrote: > On Jun 29, 4:18pm, Garance A Drosihn wrote: > } Subject: Re: PPro vs PII > > > > It also doesn't address price/performance considerations. I went > > with a dual PPro system simply because I stumbled upon a great > > deal. On the other hand, I got that great deal because PPro's > > any left? Lots and lots. www.pricewatch.com will point you at plenty. Intel Providence board with integrated 2940UW, 10/100 EtherExpress and sound ~= $99. Various places, or www.compgeeks.com. Micron W6-Li in various configurations from $59 to $150 (www.onsale.com have a load). 128Mb ECC EDO DIMM (buffered, to suit 440FX) ~= $170. PPro 180/256 ~= $99 each. PPro 166/512 ~= $145 each. Bought my 166s new, OEM from Micro-Xpress, running perfectly happily @ 233Mhz :=) HTH, AS To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 03:15:14 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id DAA18387 for freebsd-smp-outgoing; Tue, 30 Jun 1998 03:15:14 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from bsd.synx.com (rt.synx.com [194.167.81.239]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id DAA18327 for ; Tue, 30 Jun 1998 03:15:02 -0700 (PDT) (envelope-from <@rn.synx.com:root@synx.com>) Received: from s3.synx.com (s3 [192.1.1.247]) by bsd.synx.com (8.6.12/8.6.12) with SMTP id LAA22529; Tue, 30 Jun 1998 11:14:22 +0100 Received: from rn by s3.synx.com id aa09614; 30 Jun 98 12:02 BST Date: Tue, 30 Jun 1998 12:14:11 +0200 (CEST) From: Remy NONNENMACHER Reply-To: remy@synx.com Subject: Re: PPro vs PII To: spadger@best.com cc: bryan@coombs.anu.edu.au, seth@pengar.com, freebsd-smp@FreeBSD.ORG In-Reply-To: <35989768.71FABE90@best.com> MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Message-ID: <9806301202.aa09614@s3.synx.com> Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 30 Jun, Andy Sparrow wrote: > bryan collins wrote: > >> Seth Leigh said: >> > Intel Providence (PR440FX) mobo, one 64 MB DIMM >> > Two Pentium Pro 180/256 overclocked (solid as rocks) to 200 MHz >> > FreeBSD 3.0-031198-SNAP >> > Running two RC5 clients, each in its own xterm, each process averages >> > around 526 kKeys/sec sustained, for a combined key rate of around 1056 >> > kKeys/sec. If I don't run X Windows (or anything else for that matter) and >>...... >> I have a dual PentiumII 350mhz on a SuperMicro P6DBS motherboard. >> >> I get 1.1million keys/sec with DES >> and 654k keys/sec with RC5 >> >..... > Micron W6-Li, (440FX "Natoma" chipset with integrated 2940UW & sound), > fitted with dual P6 166/512s overclocked to 233Mhz, running -current from > Jun 14th, 128Mb, 4Gb Seagate. > > Benchmarks at 1M key/sec with DES, 652k keys/sec with RC5, whilst running > X, Netscrape, GIMP. > Well, I was silly proposing using rc5des. As i forgot to mention that numbers can be compared only on *SAME* version. (for exemple, my P2/300 WS give me 832kk/s and the big central server here (2xPII 337i, SMP from December) give 2x925kk/s. As we can see, all these numbers are un-matchable to those given above). Shame on me and let's stop annoying the list. For those interested, send *ME* your numbers and i will post a synthetic table. Have a nice day. RN. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 06:24:46 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA19132 for freebsd-smp-outgoing; Tue, 30 Jun 1998 06:24:46 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from wrath.cs.utah.edu (wrath.cs.utah.edu [155.99.198.100]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA19112; Tue, 30 Jun 1998 06:24:37 -0700 (PDT) (envelope-from vanmaren@fast.cs.utah.edu) Received: from fast.cs.utah.edu (fast.cs.utah.edu [155.99.212.1]) by wrath.cs.utah.edu (8.8.8/8.8.8) with SMTP id HAA15540; Tue, 30 Jun 1998 07:24:36 -0600 (MDT) Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id HAA01216; Tue, 30 Jun 1998 07:24:35 -0600 Date: Tue, 30 Jun 1998 07:24:35 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Message-Id: <199806301324.HAA01216@fast.cs.utah.edu> To: freebsd@atipa.com, leo@talcom.net Subject: Re: Motherboard for Xeon Processor Cc: freebsd-smp@FreeBSD.ORG, sos@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > The new chip needs a new chipset and motherboard (Slot 2 SEC). Here are > the details of the first board to support the new architecture: Also try these URLs: http://www.ami.com/motherboards/motherboard.html http://www.ami.com/motherboards/p2/s769spec.html (I found that on Saturday.) and: http://www.dell.com/products/poweredge/pe6300/index.htm http://www.dell.com/products/poweredge/pe6300/spec.htm Note that both these use on 64-bit, 33MHz PCI bus, and 2 32-bit, 33MHz busses: the 450GX has 2 64-bit PCI controllers that can be configured as a single 64-bit PCI or two 32-bit PCI, so there are other options. Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 06:47:32 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id GAA23300 for freebsd-smp-outgoing; Tue, 30 Jun 1998 06:47:32 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from indigo.ie (nsmart@ts01-02.waterford.indigo.ie [194.125.139.65]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id GAA23265 for ; Tue, 30 Jun 1998 06:47:19 -0700 (PDT) (envelope-from rotel@indigo.ie) Received: (from nsmart@localhost) by indigo.ie (8.8.8/8.8.7) id OAA00501; Tue, 30 Jun 1998 14:43:06 +0100 (IST) (envelope-from rotel@indigo.ie) From: Niall Smart Message-Id: <199806301343.OAA00501@indigo.ie> Date: Tue, 30 Jun 1998 14:43:05 +0000 In-Reply-To: Andy Sparrow "Re: PPro vs PII" (Jun 30, 12:58am) Reply-To: rotel@indigo.ie X-Mailer: Mail User's Shell (7.2.6 beta(3) 11/17/96) To: Andy Sparrow , rotel@indigo.ie Subject: Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Jun 30, 12:58am, Andy Sparrow wrote: } Subject: Re: PPro vs PII > Niall Smart wrote: > > > On Jun 29, 4:18pm, Garance A Drosihn wrote: > > } Subject: Re: PPro vs PII > > > > > > It also doesn't address price/performance considerations. I went > > > with a dual PPro system simply because I stumbled upon a great > > > deal. On the other hand, I got that great deal because PPro's > > > > any left? > > Lots and lots. Hrm, I didn't realise I'd forwarded this to the list. > PPro 180/256 ~= $99 each. > > PPro 166/512 ~= $145 each. > > Bought my 166s new, OEM from Micro-Xpress, running perfectly happily > @ 233Mhz :=) Well, this is the big decision, 166->233 sounds very ambitious, and I don't want a computer that blows things around my room. :) I'm thinking 2xP200/256 running at 75 * 3, (not sure if thats possible with the LX) or 66 * 3.5. Surely the 166/512's were the ones that wouldn't run reliably at 200? How long does it take you to buildworld? Niall -- Niall Smart. PGP: finger njs3@motmot.doc.ic.ac.uk FreeBSD: Turning PC's into Workstations: www.freebsd.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 07:17:13 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id HAA28042 for freebsd-smp-outgoing; Tue, 30 Jun 1998 07:17:13 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from wrath.cs.utah.edu (wrath.cs.utah.edu [155.99.198.100]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id HAA28026 for ; Tue, 30 Jun 1998 07:17:05 -0700 (PDT) (envelope-from vanmaren@fast.cs.utah.edu) Received: from fast.cs.utah.edu (fast.cs.utah.edu [155.99.212.1]) by wrath.cs.utah.edu (8.8.8/8.8.8) with SMTP id IAA16446 for ; Tue, 30 Jun 1998 08:17:02 -0600 (MDT) Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id IAA01537; Tue, 30 Jun 1998 08:17:01 -0600 Date: Tue, 30 Jun 1998 08:17:01 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Message-Id: <199806301417.IAA01537@fast.cs.utah.edu> To: freebsd-smp@FreeBSD.ORG Subject: Re: Motherboard for Xeon Processor Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > http://developer.intel.com/design/wrkstn/MS440gx/index.htm It has integrated 10/100 ethernet and 6 PCI slots -- all on a single PCI bus (I checked the PCI configuration space map to be sure). Intel must be doing something different from everyone else to get that many electrical loads on a single PCI bus. [I fully expected a PCI-PCI bridge, like on some Tyan and AMI motherboards.] It is either get the 440GX chipset, which has a single PCI bus, but supports 2 GB SDRAM, or get the 450NX, which supports 2-4 PCI busses, and uses 50ns EDO DIMMS, with 4-way interleaving. One more MB: http://www.tyan.com/html/body_slot2.html To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 09:05:01 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA14697 for freebsd-smp-outgoing; Tue, 30 Jun 1998 09:05:01 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA14672 for ; Tue, 30 Jun 1998 09:04:47 -0700 (PDT) (envelope-from root@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id JAA14782; Tue, 30 Jun 1998 09:04:29 -0700 (PDT) Message-Id: <199806301604.JAA14782@implode.root.com> To: vanmaren@fast.cs.utah.edu (Kevin Van Maren) cc: freebsd-smp@FreeBSD.ORG Subject: Re: Motherboard for Xeon Processor In-reply-to: Your message of "Tue, 30 Jun 1998 08:17:01 MDT." <199806301417.IAA01537@fast.cs.utah.edu> From: David Greenman Reply-To: dg@root.com Date: Tue, 30 Jun 1998 09:04:29 -0700 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > http://developer.intel.com/design/wrkstn/MS440gx/index.htm I noticed that the above has only 4 DIMM sockets yet they claim up to 2GB of SDRAM, so I'm curious when the 512MB SDRAM DIMMs will be available. -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 10:31:22 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id KAA27635 for freebsd-smp-outgoing; Tue, 30 Jun 1998 10:31:22 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from proxy4.ba.best.com (root@proxy4.ba.best.com [206.184.139.15]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id KAA27614 for ; Tue, 30 Jun 1998 10:31:15 -0700 (PDT) (envelope-from spadger@best.com) Received: from best.com (dynamic44.pm05.sf3d.best.com [209.24.235.44]) by proxy4.ba.best.com (8.8.8/8.8.BEST) with ESMTP id KAA26060; Tue, 30 Jun 1998 10:25:59 -0700 (PDT) Message-ID: <35991FA6.411F3396@best.com> Date: Tue, 30 Jun 1998 10:25:58 -0700 From: Andy Sparrow Organization: Not likely. X-Mailer: Mozilla 4.05 [en] (X11; I; FreeBSD 3.0-CURRENT i386) MIME-Version: 1.0 To: rotel@indigo.ie CC: freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII References: <199806301343.OAA00501@indigo.ie> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Niall Smart wrote: > Hrm, I didn't realise I'd forwarded this to the list. O well :-) > > PPro 180/256 ~= $99 each. > > > > PPro 166/512 ~= $145 each. > > > > Bought my 166s new, OEM from Micro-Xpress, running perfectly happily > > @ 233Mhz :=) > > Well, this is the big decision, 166->233 sounds very ambitious, and I > don't want a computer that blows things around my room. :) Neither do I. And a computer that flakes out when you're trying touse it is utterly pointless. That said, they've been running @ 233 Mhz, 24/7 for months. No crashes, no lockups. However, they're NOT running stupid little $10 fans, they're running an efficient (big) heatsink and 20CFM fans. With thermal compound. And an extra case fan, in a big case. They actually run considerably cooler @ 233 than the 200Mhz P6 in a standard Micron (my machine at work is stock, and border-line thermally unstable due to an inferior fan-and-no-heatsink design - it won't boot @ 233, and crashes sometimes @ 200 {A/C not very effective in my office}). Now, in this instance, /266/ is too ambitious for my home machine - they won't load the FreeBSD kernel at that speed :=) > I'm thinking > 2xP200/256 running at 75 * 3, (not sure if thats possible with the LX) Intel chipsets don't officially support any bus speed higher than 66Mhz until the BX, IIRC.Read the (-smp) archives. Some time ago, someone asked whether to gofor 180/256 or 166/512 for SMP. They were advised to get the largest L2 cache possible for SMP, as this would more than offset the effect of the slower clock speed. That advice arrived in the nick of time for me - at the time, the 180s were nearly as expensive as the 166s, only $10 cheaper. And I knew that it would be a cruel fate indeed that wouldn't let me take a 'GenuineIntel' 166Mhz CPU to 180.. > or 66 * 3.5. Surely the 166/512's were the ones that wouldn't run > reliably at 200? Obviously not...They ran perfectly happily @ 200 as well :=) Interestingly, the CPU id from my 166s is exactly the same as the one from my 200, from dmesg: /kernel: CPU: Pentium Pro (686-class CPU) /kernel: Origin = "GenuineIntel" Id = 0x619 Stepping=9 > How long does it take you to buildworld? Just over 90 minutes. NOPROFILE=true, CFLAGS="-O -pipe", with -j24, target filesystem mounted with noatime, async The Seagate Hawk 4 (SCA->Fast Wide) drive used for both the source and the object files is probably the reason that this isn't faster. But a brand new 4Gb drive for $139? I couldn't resist it :=) Anyway, a discussion about overclocking is getting a litle off-topic for SMP, so, whilst I'd be happy to discuss it with anyone interested via email, let's trim the cc line? Cheers, AS To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 11:12:08 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA04738 for freebsd-smp-outgoing; Tue, 30 Jun 1998 11:12:08 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail1.its.rpi.edu (root@mail1.its.rpi.edu [128.113.100.7]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA04669; Tue, 30 Jun 1998 11:11:57 -0700 (PDT) (envelope-from drosih@rpi.edu) Received: from [128.113.24.47] (gilead.acs.rpi.edu [128.113.24.47]) by mail1.its.rpi.edu (8.8.8/8.8.6) with ESMTP id OAA07864; Tue, 30 Jun 1998 14:11:21 -0400 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Sender: drosih@pop1.rpi.edu Message-Id: In-Reply-To: References: <19980629200753.14325@supersex.com> Date: Tue, 30 Jun 1998 14:15:09 -0400 To: Atipa From: Garance A Drosihn Subject: Re: Intel announces Xeon!!! Was Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG, sos@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org At 8:47 PM -0600 6/29/98, Atipa wrote: > This thread is now obselete, since the new Xeon has the best features > of both chips. People checking out URL's might also want to look at: http://www.edtn.com/news/june26/062698tnews1.html which is a news article about a bug in the Xeon. They (Intel) aren't specific about what the bug was, but apparently it only effects the use of Xeon in "servers", but not "workstations". They also have a microcode fix which they are testing. The upshot of all this is that "server configurations" based on the Xeon may be delayed by a few weeks. --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer or drosih@rpi.edu Rensselaer Polytechnic Institute To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 11:20:18 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA06713 for freebsd-smp-outgoing; Tue, 30 Jun 1998 11:20:18 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA06642 for ; Tue, 30 Jun 1998 11:20:10 -0700 (PDT) (envelope-from jhay@zibbi.mikom.csir.co.za) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.9.0/8.9.0) id UAA19361; Tue, 30 Jun 1998 20:15:28 +0200 (SAT) From: John Hay Message-Id: <199806301815.UAA19361@zibbi.mikom.csir.co.za> Subject: Re: time problem? In-Reply-To: <9570.899103386@critter.freebsd.dk> from Poul-Henning Kamp at "Jun 29, 98 08:56:26 am" To: phk@critter.freebsd.dk (Poul-Henning Kamp) Date: Tue, 30 Jun 1998 20:15:28 +0200 (SAT) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > >---------------- > >A: 899000790.981073054 #29011 D: 1.000004191 > >A: 899000792.826104370 #29012 D: 1.845031316 > >A: 899016109.813482765 #44329 D: 0.999940488 > >A: 899016111.668519664 #44330 D: 1.855036899 > >---------- > > > >So on sequence numbers 29012 and 44330, I have gained .85 seconds > >in the space of a second. For the 15000 seconds between 29012 and > >44329 there were no glitches or jumps. > > Now, one source of this could be interrupts diasabled for too long, > since your timestamps depend on the sio interrupt that would be my > primary suspect. > I would have thought that disabling interrupts for too long would cause the machine to loose time not to gain it? I followed Tony Li's advice and started ntpd again on the machine with no local reference clocks and the machine has now been up for more than 24 hours with no jumps at all, so it looks like it is calling nanotime() from within the sio interrupt that causes the problems. :-( So what are my options then? Can I somehow run the sio interrupts at a lower priority? I would really prefer to have the pps signal on the serial port, because the port is already used to read the serial stream from the GPS. John -- John Hay -- John.Hay@mikom.csir.co.za To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 11:28:52 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA08572 for freebsd-smp-outgoing; Tue, 30 Jun 1998 11:28:52 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.133.1] (may be forged)) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA08451 for ; Tue, 30 Jun 1998 11:28:22 -0700 (PDT) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.5) with ESMTP id UAA15033; Tue, 30 Jun 1998 20:23:02 +0200 (CEST) To: John Hay cc: freebsd-smp@FreeBSD.ORG Subject: Re: time problem? In-reply-to: Your message of "Sat, 30 Jun 1998 20:15:28 +0200." <199806301815.UAA19361@zibbi.mikom.csir.co.za> Date: Tue, 30 Jun 1998 20:23:02 +0200 Message-ID: <15031.899230982@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >I followed Tony Li's advice and started ntpd again on the machine with >no local reference clocks and the machine has now been up for more than >24 hours with no jumps at all, so it looks like it is calling nanotime() >from within the sio interrupt that causes the problems. :-( yes, that may not be entirely safe to do according to Bruce. >So what are my options then? Can I somehow run the sio interrupts at >a lower priority? I would really prefer to have the pps signal on the >serial port, because the port is already used to read the serial stream >from the GPS. I don't know :-) You're entering territory here where I have not yet managed to venture... I pressume you havn't used the PPS_SYNC stuff in the kernel (if not: don't!) And it doesn't happen if you run !SMP with the TSC timecounter, right ? -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." "ttyv0" -- What UNIX calls a $20K state-of-the-art, 3D, hi-res color terminal To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 11:51:28 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA12389 for freebsd-smp-outgoing; Tue, 30 Jun 1998 11:51:28 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from wrath.cs.utah.edu (wrath.cs.utah.edu [155.99.198.100]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA12366; Tue, 30 Jun 1998 11:51:18 -0700 (PDT) (envelope-from vanmaren@fast.cs.utah.edu) Received: from fast.cs.utah.edu (fast.cs.utah.edu [155.99.212.1]) by wrath.cs.utah.edu (8.8.8/8.8.8) with SMTP id MAA23369; Tue, 30 Jun 1998 12:51:01 -0600 (MDT) Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id MAA04665; Tue, 30 Jun 1998 12:51:01 -0600 Date: Tue, 30 Jun 1998 12:51:01 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Message-Id: <199806301851.MAA04665@fast.cs.utah.edu> To: freebsd@atipa.com, ken@plutotech.com Subject: Re: Intel announces Xeon!!! Was Re: PPro vs PII Cc: freebsd-smp@FreeBSD.ORG, leo@talcom.net, sos@FreeBSD.ORG Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org wysiwyg://30/http://developer.intel.com/design/servers/ "shipping 3rd quarter" To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 11:55:25 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA12917 for freebsd-smp-outgoing; Tue, 30 Jun 1998 11:55:25 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id LAA12895 for ; Tue, 30 Jun 1998 11:55:19 -0700 (PDT) (envelope-from jhay@zibbi.mikom.csir.co.za) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.9.0/8.9.0) id UAA19830; Tue, 30 Jun 1998 20:47:42 +0200 (SAT) From: John Hay Message-Id: <199806301847.UAA19830@zibbi.mikom.csir.co.za> Subject: Re: time problem? In-Reply-To: <15031.899230982@critter.freebsd.dk> from Poul-Henning Kamp at "Jun 30, 98 08:23:02 pm" To: phk@critter.freebsd.dk (Poul-Henning Kamp) Date: Tue, 30 Jun 1998 20:47:42 +0200 (SAT) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL32 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org > > >I followed Tony Li's advice and started ntpd again on the machine with > >no local reference clocks and the machine has now been up for more than > >24 hours with no jumps at all, so it looks like it is calling nanotime() > >from within the sio interrupt that causes the problems. :-( > > yes, that may not be entirely safe to do according to Bruce. > > >So what are my options then? Can I somehow run the sio interrupts at > >a lower priority? I would really prefer to have the pps signal on the > >serial port, because the port is already used to read the serial stream > >from the GPS. > > I don't know :-) Hmmm, so it seems that I'll have to get the soldering out and get the pps signal hooked up to the printer port to see if that works better. :-) > > You're entering territory here where I have not yet managed to venture... > > I pressume you havn't used the PPS_SYNC stuff in the kernel (if not: don't!) Well I have and on the old 486 it works pretty ok, so I guess I shouldn't upgrade it until I have resolved this. :-) What I do know is the way you calculate the second pararameter for hardpps() in your printer port pps driver and the way I did it in my patch, seems to make it very unstable. :-) Here is how it looks on the 486: ------------- # /usr/local/sbin/ntpdc shark ntpdc> kerni pll offset: 0 us pll frequency: -14.932 ppm maximum error: 15266 us estimated error: 1 us status: 0107 pll time constant: 2 precision: 1 us frequency tolerance: 512 ppm pps frequency: -55.983 ppm pps stability: 0.010 ppm pps jitter: 2 us calibration interval: 256 s calibration cycles: 6790 jitter exceeded: 450 stability exceeded: 4 calibration errors: 37 ntpdc> peer remote local st poll reach delay offset disp ======================================================================= =GPS_NMEA(0) 0.0.0.0 0 64 377 0.00000 -0.000049 0.00092 *PPS(0) 0.0.0.0 0 64 377 0.00000 0.000001 0.00092 +orca.cids.org.z 0.0.0.0 1 64 0 0.00000 0.000000 16.0000 ---------------- > And it doesn't happen if you run !SMP with the TSC timecounter, right ? I'll have a look at it again. John -- John Hay -- John.Hay@mikom.csir.co.za To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 12:19:07 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id MAA15981 for freebsd-smp-outgoing; Tue, 30 Jun 1998 12:19:07 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from critter.freebsd.dk (critter.freebsd.dk [195.8.133.1] (may be forged)) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id MAA15971 for ; Tue, 30 Jun 1998 12:18:58 -0700 (PDT) (envelope-from phk@critter.freebsd.dk) Received: from critter.freebsd.dk (localhost [127.0.0.1]) by critter.freebsd.dk (8.8.7/8.8.5) with ESMTP id VAA15174; Tue, 30 Jun 1998 21:13:34 +0200 (CEST) To: John Hay cc: freebsd-smp@FreeBSD.ORG Subject: Re: time problem? In-reply-to: Your message of "Sat, 30 Jun 1998 20:47:42 +0200." <199806301847.UAA19830@zibbi.mikom.csir.co.za> Date: Tue, 30 Jun 1998 21:13:33 +0200 Message-ID: <15172.899234013@critter.freebsd.dk> From: Poul-Henning Kamp Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org >> >I followed Tony Li's advice and started ntpd again on the machine with >> >no local reference clocks and the machine has now been up for more than >> >24 hours with no jumps at all, so it looks like it is calling nanotime() >> >from within the sio interrupt that causes the problems. :-( >> >> yes, that may not be entirely safe to do according to Bruce. >> >> >So what are my options then? Can I somehow run the sio interrupts at >> >a lower priority? I would really prefer to have the pps signal on the >> >serial port, because the port is already used to read the serial stream >> >from the GPS. >> >> I don't know :-) > >Hmmm, so it seems that I'll have to get the soldering out and get the >pps signal hooked up to the printer port to see if that works better. :-) He's not too happy about that either... >What I do know is the way you calculate the second pararameter for >hardpps() in your printer port pps driver and the way I did it in >my patch, seems to make it very unstable. :-) yes, that parameter is a kludge for the hardclock() interpolation stuff which Dave Mills wrote for alphas. >> And it doesn't happen if you run !SMP with the TSC timecounter, right ? > >I'll have a look at it again. Please do, because the TSC doesn't require the interrupts to be disabled, so that would be a very valuable datapoint for us... -- Poul-Henning Kamp FreeBSD coreteam member phk@FreeBSD.ORG "Real hackers run -current on their laptop." "ttyv0" -- What UNIX calls a $20K state-of-the-art, 3D, hi-res color terminal To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 18:06:22 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA18473 for freebsd-smp-outgoing; Tue, 30 Jun 1998 18:06:22 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from electric.tbe.net (electric.tbe.net [207.99.115.10]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id SAA18379 for ; Tue, 30 Jun 1998 18:06:00 -0700 (PDT) (envelope-from gary@tbe.net) Received: (qmail 11397 invoked from network); 1 Jul 1998 01:04:39 -0000 Received: from electric.tbe.net (207.99.115.10) by electric.tbe.net with SMTP; 1 Jul 1998 01:04:39 -0000 Date: Tue, 30 Jun 1998 21:04:37 -0400 (EDT) From: "Gary D. Margiotta" To: Andy Sparrow cc: rotel@indigo.ie, freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: <35989A91.BBF90786@best.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org In the case of PPro's, with a price difference of ~$50 per chip, does it pay to spend the extra on the 166/512k ($145) as compared to the 180/256k? ($95) Straight clock speeds should dictate that the 166/512k would be a little faster due to the larger cache onboard. But if you overclock both to, say, 200, is there enough of a difference to notice it? I can study every benchmark and come out with a winner on paper, but is the price/performance worth it... ______________________________________________________________ -Gary Margiotta Voice: (973) 835-9696 TBE Internet Services Fax: (973) 835-2133 http://www.tbe.net E-Mail: gary@tbe.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Tue Jun 30 18:09:45 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA19108 for freebsd-smp-outgoing; Tue, 30 Jun 1998 18:09:45 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail1.its.rpi.edu (root@mail1.its.rpi.edu [128.113.100.7]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA19038 for ; Tue, 30 Jun 1998 18:09:16 -0700 (PDT) (envelope-from drosih@rpi.edu) Received: from [128.113.24.47] (gilead.acs.rpi.edu [128.113.24.47]) by mail1.its.rpi.edu (8.8.8/8.8.6) with ESMTP id VAA104164; Tue, 30 Jun 1998 21:09:03 -0400 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-Sender: drosih@pop1.rpi.edu Message-Id: In-Reply-To: Date: Tue, 30 Jun 1998 21:12:52 -0400 To: Seth Leigh , freebsd-smp@FreeBSD.ORG From: Garance A Drosihn Subject: Re: Temp. Monitoring Software for Intel Providence mobo? Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org At 8:19 PM +0100 6/15/98, Seth Leigh wrote: >Hi all. > I know a few people have bought Intel PR440FX Providence dual PPro > motherboards recently, as have I. My machine is up and running just > great now. I want to be able to monitor the temperatures of my two > PPro cpus. Does anyone know if there is any software available already > that can talk to the motherboard's temp. sensing hardware? Some other articles followed this, answering some other questions, but I don't think I ever saw an answer to this. Was there any software for monitoring the motherboard temperatures? (I have the same motherboard in two machines, with one of them setup with two PPro's...) --- Garance Alistair Drosehn = gad@eclipse.its.rpi.edu Senior Systems Programmer or drosih@rpi.edu Rensselaer Polytechnic Institute To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Wed Jul 1 13:15:21 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA13912 for freebsd-smp-outgoing; Wed, 1 Jul 1998 13:15:21 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from neunacht.netgsi.com (neunacht.netgsi.com [192.55.203.174]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA13892 for ; Wed, 1 Jul 1998 13:15:15 -0700 (PDT) (envelope-from cjohnson@neunacht.netgsi.com) Received: (from cjohnson@localhost) by neunacht.netgsi.com (8.8.8/8.8.6) id QAA01480 for freebsd-smp@freebsd.org; Wed, 1 Jul 1998 16:15:11 -0400 (EDT) From: "Christopher T. Johnson" Message-Id: <199807012015.QAA01480@neunacht.netgsi.com> Subject: rfork() questions To: freebsd-smp@FreeBSD.ORG Date: Wed, 1 Jul 1998 16:15:10 -0400 (EDT) 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 Precedence: bulk X-Loop: FreeBSD.org Well, I have this very nice raytracing code, it has been doing parallel processing for many years now. The only thing it needs is the ability to do a fork() while sharing memory, which rfork(RFPROC|RFMEM|RFFDG) should do nicely, and semaphores. We've looked and found the semaphore syscall. And rfork() says it should work, but of course rfork() fails if you are in SMP mode and request rfork() with the RFMEM flag. I and my friends have no problem making kernel mods, we've been doing it for a number of years. We are looking for a pointer to whomever is working on the rfork() shared DPT problems. Thanks in advance for any help or information. Chris Johnson To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Wed Jul 1 22:10:57 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA02681 for freebsd-smp-outgoing; Wed, 1 Jul 1998 22:10:57 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from hobbes.saturn-tech.com (drussell@drussell.internode.net [198.161.228.154]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA02676 for ; Wed, 1 Jul 1998 22:10:55 -0700 (PDT) (envelope-from drussell@saturn-tech.com) Received: from localhost (drussell@localhost) by hobbes.saturn-tech.com (8.8.4/8.8.2) with SMTP id XAA27024; Wed, 1 Jul 1998 23:11:42 -0600 (MDT) Date: Wed, 1 Jul 1998 23:11:42 -0600 (MDT) From: Doug Russell To: "Jordan K. Hubbard" cc: Atipa , freebsd-smp@FreeBSD.ORG Subject: Re: PPro vs PII In-Reply-To: <7197.899192320@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, 30 Jun 1998, Jordan K. Hubbard wrote: > P.S. You never did take up my challenge of proposing a better ports > collection design after raising such a ruckus about it before. Do you > also design, or only whine? :) What are we talking about, the directory structure? (Basically?) Later...... To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Thu Jul 2 09:24:22 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA22281 for freebsd-smp-outgoing; Thu, 2 Jul 1998 09:24:22 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.rdc1.on.wave.home.com (mta@ha1.rdc1.on.wave.home.com [24.2.9.66]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id JAA22223 for ; Thu, 2 Jul 1998 09:24:12 -0700 (PDT) (envelope-from dgrant1@rogers.wave.ca) Received: from cr267111-a (cr267111-a.pr1.on.wave.home.com [24.112.97.58]) by mail.rdc1.on.wave.home.com (Netscape Messaging Server 3.5) with SMTP id AAA329B; Thu, 2 Jul 1998 09:24:06 -0700 Message-ID: <359BB4C4.4164@rogers.wave.ca> Date: Thu, 02 Jul 1998 12:26:44 -0400 From: David Grant Organization: Grant Computers X-Mailer: Mozilla 3.0 (Win95; U) MIME-Version: 1.0 To: freebsd-smp@FreeBSD.ORG CC: motheraforever@yahoo.com Subject: cool Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org this is cool it rocks To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Thu Jul 2 17:04:00 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA02283 for freebsd-smp-outgoing; Thu, 2 Jul 1998 17:04:00 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from mail.atipa.com (altrox.atipa.com [208.128.22.34]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id RAA02270 for ; Thu, 2 Jul 1998 17:03:56 -0700 (PDT) (envelope-from kmcc@atipa.com) Received: (qmail 1060 invoked by uid 500); 2 Jul 1998 23:00:56 -0000 Date: Thu, 2 Jul 1998 17:00:56 -0600 (MDT) From: Kevin McCarthy To: smp@FreeBSD.ORG Subject: Product Discontinuance Message (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org ---------- Forwarded message ---------- Date: Thu, 2 Jul 1998 16:36:23 -0700 (PDT) From: Intel To: Intel Customers Subject: Product Discontinuance Message ---------------------------------------------------- Product Discontinuance on all Pentium(R) Pro Processors 166-,180- and 200-MHz ---------------------------------------------------- July 1, 1998 Effective July 1st, 1998, the boxed Pentium(R) Pro processors 166-, 180 and 200-MHz are formally discontinued. Industry demand for Pentium Pro processors has shifted to the Intel(R) Pentium II and Pentium II Xeon(TM) processors. Frequency Cache Package Box Order Code 166 MHz 512k SPGA BOXBP80521166512 180 MHz 256k SPGA (A,E) BOXBP80521180 Bulk 180 MHz 256k SPGA BULKBP80521180 200 MHz 256k SPGA (A,E) BOXBP80521200 Bulk 200 MHz 256k SPGA BULKBP80521200 This Product Discontinuance does not affect or modify any warranty or technical support obligation set forth in the terms and conditions governing your original purchases of the boxed processors. Specifically, Intel boxed processors are covered by a three year limited warranty, and Intel will provide standard warranty and technical support for all boxed processors for the duration of the warranty period. If you have any questions or concerns, please contact your Intel representative. A list of representatives is available at I.P.D. Net at: http://channel.intel.com/business/america/private/ipd2/support/nachamp.htm generic user ID: ipd6 password: solutions ---------------------------------------------------- Please do not respond directly to this message. If you would like further information or have feedback, just let us know at: ipdmail@cs.intel.com or (888)617-7884. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 14:10:35 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA18944 for freebsd-smp-outgoing; Fri, 3 Jul 1998 14:10:35 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from marlin.corp.gulf.net (root@marlin.corp.gulf.net [198.69.72.17]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA18926; Fri, 3 Jul 1998 14:10:32 -0700 (PDT) (envelope-from tbackman@corp.gulf.net) Received: from marlin.corp.gulf.net (tbackman@marlin.corp.gulf.net [206.105.61.2]) by marlin.corp.gulf.net (8.8.8/8.8.5) with SMTP id QAA11835; Fri, 3 Jul 1998 16:09:37 -0500 (CDT) Date: Fri, 3 Jul 1998 16:09:37 -0500 (CDT) From: Todd Backman To: Kyle Mestery cc: freebsd-smp@FreeBSD.ORG, freebsd-hardware@FreeBSD.ORG Subject: SMP FreeBSD Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I am waiting for UPS to deliver my BUFFERED 64mb dimms at this moment. Any suggestions on the SMP install? I have dl'd the 3.0 boot.flp and am curious as to how I incorporate CAM with this. I bought the Micronics Dual PPRO (W6-LI-007) for $54 from onsale.com and am ready to start on my little "project" for the weekend (my wife just can't understand this madness); ^) . Any suggestions would be great. Thanks in advance. ===================================================================== Todd Backman (tbackman@corp.gulf.net) Network Engineering Team Leader Systems/POP Administration Gulf Coast Internet Company 1-800-444-INET To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 14:58:02 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id OAA23003 for freebsd-smp-outgoing; Fri, 3 Jul 1998 14:58:02 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from icicle.winternet.com (adm@icicle.winternet.com [198.174.169.13]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id OAA22965 for ; Fri, 3 Jul 1998 14:57:59 -0700 (PDT) (envelope-from mestery@mail.winternet.com) Received: (from adm@localhost) by icicle.winternet.com (8.8.8/8.8.8) id QAA22330; Fri, 3 Jul 1998 16:57:56 -0500 (CDT) Received: from tundra.winternet.com(198.174.169.11) by icicle.winternet.com via smap (V2.0) id xma022311; Fri, 3 Jul 98 16:57:45 -0500 Received: from localhost (mestery@localhost) by tundra.winternet.com (8.8.7/8.8.4) with SMTP id QAA26794; Fri, 3 Jul 1998 16:57:42 -0500 (CDT) X-Authentication-Warning: tundra.winternet.com: mestery owned process doing -bs Date: Fri, 3 Jul 1998 16:57:42 -0500 (CDT) From: Kyle Mestery To: Todd Backman cc: freebsd-smp@FreeBSD.ORG Subject: Re: SMP FreeBSD In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, 3 Jul 1998, Todd Backman wrote: > > I am waiting for UPS to deliver my BUFFERED 64mb dimms at this moment. Any > suggestions on the SMP install? I have dl'd the 3.0 boot.flp and am > curious as to how I incorporate CAM with this. I bought the Micronics Dual > PPRO (W6-LI-007) for $54 from onsale.com and am ready to start on my > little "project" for the weekend (my wife just can't understand this > madness); ^) . Any suggestions would be great. > One thing to watch out for with this board is that you partition the disk so it is compatible with DOS style partition tables. Otherwise it will refuse to boot (at least from a SCSI disk). Mike Smith had said this was a problem with the BIOS or sometjhing, but the board works fine when you do it this way. It is a screamer. If you have decent fans, you shouild easily be able to overclock this board. -- Kyle Mestery StorageTek's Network Systems Group "I'll take what you're willing to give, and I'll teach myself to live, with a walk-on part of a background shot from a movie I'm not in." - Blink 182, "Apple Shampoo" To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 15:00:18 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id PAA23240 for freebsd-smp-outgoing; Fri, 3 Jul 1998 15:00:18 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from panzer.plutotech.com (ken@panzer.plutotech.com [206.168.67.125]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id PAA23221; Fri, 3 Jul 1998 15:00:13 -0700 (PDT) (envelope-from ken@panzer.plutotech.com) Received: (from ken@localhost) by panzer.plutotech.com (8.8.8/8.8.5) id QAA20397; Fri, 3 Jul 1998 16:00:02 -0600 (MDT) From: "Kenneth D. Merry" Message-Id: <199807032200.QAA20397@panzer.plutotech.com> Subject: Re: SMP FreeBSD In-Reply-To: from Todd Backman at "Jul 3, 98 04:09:37 pm" To: tbackman@corp.gulf.net (Todd Backman) Date: Fri, 3 Jul 1998 16:00:02 -0600 (MDT) Cc: mestery@winternet.com, freebsd-smp@FreeBSD.ORG, freebsd-hardware@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 Precedence: bulk X-Loop: FreeBSD.org Todd Backman wrote... > > I am waiting for UPS to deliver my BUFFERED 64mb dimms at this moment. Any > suggestions on the SMP install? I have dl'd the 3.0 boot.flp and am > curious as to how I incorporate CAM with this. I bought the Micronics Dual > PPRO (W6-LI-007) for $54 from onsale.com and am ready to start on my > little "project" for the weekend (my wife just can't understand this > madness); ^) . Any suggestions would be great. To incorporate CAM, follow the instructions in the readme: ftp://ftp.FreeBSD.ORG/pub/FreeBSD/cam/README or ftp://ftp.kdm.org/pub/FreeBSD/cam/README The patches that are out there were made May 20th, so you'll probably want a snapshot from somewhere around that vintage from current.freebsd.org. We're working on a new CAM snapshot, with the aim of having it out sometime this weekend. (can't make any promises, but that's the plan..) Ken -- Kenneth Merry ken@plutotech.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 17:23:56 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id RAA12333 for freebsd-smp-outgoing; Fri, 3 Jul 1998 17:23:56 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from gatekeeper.falcon.com (eppp3.sysnet.net [206.142.16.84]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id RAA12302 for ; Fri, 3 Jul 1998 17:23:43 -0700 (PDT) (envelope-from patton@sysnet.net) Received: from [192.168.1.10] ([192.168.1.10]) by gatekeeper.falcon.com (8.8.8/8.8.5) with ESMTP id UAA01620 for ; Fri, 3 Jul 1998 20:15:08 -0400 (EDT) X-Sender: patton@mail.sysnet.net Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 2 Jul 1998 20:27:34 -0400 To: freebsd-smp@FreeBSD.ORG From: Matthew Patton Subject: determining SMP Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org How do I determine if my rebuilt kernel really is seeing 2 CPUs? I'm working off of 2.2.6 sources and defined options SMP options APIC_IO per http://www.freebsd.org/~fsmp/SMP/getstarted.html dmesg says: CPU: Pentium Pro (231.78-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0x619 Stepping=9 Features=0xfbff r eal memory = 134217728 (131072K bytes) avail memory = 129691648 (126652K bytes) >From reading the archives I take it that SMP support has been around for quite some time and is not limited to 3.0. Am I correct? BTW, anyone know how to get sources via CVS? I'm having a devil of a time with anoncvs.freebsd.org. The error says: % cvs -d anoncvs@anoncvs.freebsd.org:/cvs up . select: protocol failure in circuit setup cvs [update aborted]: end of file from server (consult above messages if any) -------- It is by caffeine alone I set my mind in motion, it is by the beans of Java that thoughts acquire speed, the hands acquire shaking, the shaking becomes a warning. It is by caffeine alone I set my mind in motion. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 18:18:51 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA18398 for freebsd-smp-outgoing; Fri, 3 Jul 1998 18:18:51 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from panzer.plutotech.com (ken@panzer.plutotech.com [206.168.67.125]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA18393 for ; Fri, 3 Jul 1998 18:18:47 -0700 (PDT) (envelope-from ken@panzer.plutotech.com) Received: (from ken@localhost) by panzer.plutotech.com (8.8.8/8.8.5) id TAA21461; Fri, 3 Jul 1998 19:18:40 -0600 (MDT) From: "Kenneth D. Merry" Message-Id: <199807040118.TAA21461@panzer.plutotech.com> Subject: Re: determining SMP In-Reply-To: from Matthew Patton at "Jul 2, 98 08:27:34 pm" To: patton@sysnet.net (Matthew Patton) Date: Fri, 3 Jul 1998 19:18:40 -0600 (MDT) Cc: 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: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id SAA18394 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Matthew Patton wrote... > How do I determine if my rebuilt kernel really is seeing 2 CPUs? > I'm working off of 2.2.6 sources and defined > options SMP > options APIC_IO > per http://www.freebsd.org/~fsmp/SMP/getstarted.html > > dmesg says: > CPU: Pentium Pro (231.78-MHz 686-class CPU) > Origin = "GenuineIntel" Id = 0x619 Stepping=9 > > Features=0xfbff r > eal memory = 134217728 (131072K bytes) > avail memory = 129691648 (126652K bytes) > > >From reading the archives I take it that SMP support has been around for > quite some time and is not limited to 3.0. Am I correct? You are right that it has been around for a long time. SMP is limited to 3.0-current, so your 2.2.6 kernel doesn't support SMP. You'll know SMP is working when you get something like this upon booting a SMP -current kernel: Copyright (c) 1992-1998 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Sat Jun 27 19:50:55 MDT 1998 ken@panzer.plutotech.com:/usr/home/ken/perforce/cam/sys/compile/panzer Timecounter "i8254" frequency 1193182 Hz cost 3532 ns CPU: Pentium Pro (686-class CPU) Origin = "GenuineIntel" Id = 0x617 Stepping=7 Features=0xfbff real memory = 134217728 (131072K bytes) avail memory = 127840256 (124844K 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 [ ... ] IP packet filtering initialized, divert enabled, logging limited to 100 packets/entry SMP: AP CPU #1 Launched! > BTW, anyone know how to get sources via CVS? I'm having a devil of a time > with anoncvs.freebsd.org. The error says: > % cvs -d anoncvs@anoncvs.freebsd.org:/cvs up . > select: protocol failure in circuit setup > cvs [update aborted]: end of file from server (consult above messages if any) I would suggest using cvsup, my guess is that it'll be faster than anoncvs. Ken -- Kenneth Merry ken@plutotech.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message From owner-freebsd-smp Fri Jul 3 18:40:56 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id SAA21197 for freebsd-smp-outgoing; Fri, 3 Jul 1998 18:40:56 -0700 (PDT) (envelope-from owner-freebsd-smp@FreeBSD.ORG) Received: from wrath.cs.utah.edu (wrath.cs.utah.edu [155.99.198.100]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id SAA21192 for ; Fri, 3 Jul 1998 18:40:54 -0700 (PDT) (envelope-from vanmaren@fast.cs.utah.edu) Received: from fast.cs.utah.edu (fast.cs.utah.edu [155.99.212.1]) by wrath.cs.utah.edu (8.8.8/8.8.8) with SMTP id TAA01695; Fri, 3 Jul 1998 19:40:53 -0600 (MDT) Received: by fast.cs.utah.edu (8.6.10/utah-2.15-leaf) id TAA14436; Fri, 3 Jul 1998 19:40:53 -0600 Date: Fri, 3 Jul 1998 19:40:53 -0600 From: vanmaren@fast.cs.utah.edu (Kevin Van Maren) Message-Id: <199807040140.TAA14436@fast.cs.utah.edu> To: freebsd-smp@FreeBSD.ORG, patton@sysnet.net Subject: Re: determining SMP Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org First off, SMP only works with 3.0, not the 2.2 branch. Second, to use remove CVS, set the CVSROOT environment variable to: anoncvs@anoncvs.freebsd.org:/cvs You can then do a `cvs checkout src`, or to just get the kernel, `cvs checkout src/sys`. Note that this will give you the latest 3.0 (-current) source. You can get 2.2-STABLE by specifying the 2.2 branch tag. Technically, you can run SMP on 2.2, as the original SMP work was done before 3.0 branched to be come -current, but that work was done in a separate repository. Just run 3.0-current. Anonyous cvs seems to be having problems now. You can cvsup the CVS repository, as another option. Kevin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message