From bettina at apoteelia.net Sat Mar 7 10:37:50 2009 From: bettina at apoteelia.net (Bettina Schmidtberger) Date: Sat Mar 7 10:38:10 2009 Subject: Der versprochene Geheimtipp Message-ID: <0d5f53767882768d3ae9196cd9be5b1d@localhost.localdomain> Hi Du! Wie ich es Dir versprochen habe, wollte ich Dir ja noch die Adresse sagen wo wir die Dinger bestellt haben. Gibt ja viele Seiten wo man echt nur ?bers Ohr gehauen wird. Aber bei der Adresse bekommen wir immer nur Originalware und das innheralb k?rzester Zeit zugeschickt. Mit dem Zoll hatten wir da auch nie Probleme, da der Versand direkt aus Europa erfolgt. Klasse oder? Also hier nun die Adresse: http://www.apoteelia.net Viel Spass w?nsch ich Dir und das es gut funktioniert! Gru?, Deine Bettina . . - . . . . . . . . . . : . Gib Acht! Man hatte dir eingeredet, du h?ttest es schwer, dein Leben sei verpfuscht, das Leben sei eine Schuld, sei schlecht, ohne Sinn, ohne Wert; man wollte dich ducken, dich in die gro?e Armee der Leidenden schmuggeln, du solltest bemitleidenswert werden und bemitleiden: und du glaubtest ihnen ? wie ungern! ? und wieder nicht ? wie gern! Denn du bist stark, aber warst krank ? wo? wie? was wei? ich. Und deine Sehnsucht war, herauszukommen aus allen diesen m?den Verneinungen, diesen t?richten Formeln, die im Nein ihr Ja haben, diesen t?nenden Wissenschaften, diesen Worten ?. Deswegen sprangst du von Buch zu Buch, spieltest mit ihren Formeln und lie?est sie wieder fallen, die Neins und Wenns, um selber eine zu finden, aber ein Ja! sollte sie klingen ? denn du wolltest leben! Aber nicht wie der P?bel lebt ? einen Grund, ein Ziel, eine Lebensformel suchtest du. Nun, hier ist sie: Wei?t du: das Himmelsweinglas, das du ausschl?rfen wolltest ? ? nun niete dir die Formel: Die Welt schaffst du. Du vergeistigst das Chaos zur Welt; das Andere, das Noch-nicht-Du, das alte Ding an sich, ist nur das, was von dir noch nicht geschaffen, vermenschlicht, noch nicht dein Eigentum geworden ist. ? Du schaffst die Welt: nun lebe, lebe! ? Die kleine blaue Blume l?utete so froh und stark ? warum soll ich ihr nicht glauben? Und dann bin ich baden gegangen ? ? ? und habe stundenlang im Grase gelegen; und w?hrend die wei?en Wolken durch den Himmel segelten und der Flu? geruhig durch Schilfduft und Ried und schwatzendes Vogelvolk hinstr?mte, habe ich das Ding an sich, den Intellekt und den Willen verlacht und mir ein Ich-wei?-nicht-was? gew?nscht. Gegen Abend entstiegen Schw?rme von Eintagsfliegen dem Flu?, an den Gr?sern, Halmen und Pfosten kletterten sie hoch und warfen aus der H?lle sich in die Luft zum kurzen Hochzeitsleben. Die Luft war wei? ?ber den Wassern von den auf und nieder tanzenden Massen ? und die sinkende Sonne in dem H?henrauch, den der Nordwind gebracht hatte, rot wie ein Rubin: das h?tte mich fast bezwungen, da? ich schon begann, die stundenkurze Existenz der Imago zu beklagen und daran sentimentale Folgerungen zu kn?pfen ? aber da h?rte ich den Enzian l?uten und ich lachte: Das Tier freut sich jahrelang seines R?uberlebens, und dieser Liebesflug ist sein taumelnder H?hepunkt. Es lebe das Leben und seine ewige Br?cke: Venus genetrix! Vor acht Tagen h?tte ich ihr geflucht und geklagt: Was ist das Leben? So ist das Leben: es flie?t dahin wie Wellenschaum, kommt u From robert at heron.pl Tue Mar 10 03:45:20 2009 From: robert at heron.pl (Robert Heron) Date: Tue Mar 10 03:45:28 2009 Subject: only one logical CPU used in Xeon Message-ID: <2E7DD8D9-FABA-4A30-9118-948D26DB26AE@heron.pl> Hi, I have FreeBSD 6.2R running on Intel SE7501HG2 motherboard with two Xeon installed. System detects 2 CPUs with 2 cores in each of them (total 4 cores) but it uses only one core in each Xeon (total 2 of 4 cores are used only). I see it in the 'top'. I tried the same hardware on FreeBSD 7.1R and this OS uses all of the detected cores. What should I do FreeBSD 6.2R to use all of the available cores? I cannot upgrade it to 7.x, must use 6.x Regards, Robert Heron From ady at freebsd.ady.ro Tue Mar 10 07:28:59 2009 From: ady at freebsd.ady.ro (Adrian Penisoara) Date: Tue Mar 10 07:29:06 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <2E7DD8D9-FABA-4A30-9118-948D26DB26AE@heron.pl> References: <2E7DD8D9-FABA-4A30-9118-948D26DB26AE@heron.pl> Message-ID: <78cb3d3f0903100657h77d26f04p2c464ed9a1423356@mail.gmail.com> Hi, On Tue, Mar 10, 2009 at 11:40 AM, Robert Heron wrote: > Hi, > > I have FreeBSD 6.2R running on Intel SE7501HG2 motherboard with two Xeon > installed. > System detects 2 CPUs with 2 cores in each of them (total 4 cores) but it > uses only one core in each Xeon (total 2 of 4 cores are used only). > I see it in the 'top'. I tried the same hardware on FreeBSD 7.1R and this > OS uses all of the detected cores. > > What should I do FreeBSD 6.2R to use all of the available cores? You should post your dmesg output (eventually from both 6.2 and 7.1, kernel booted verbosely with -v in bootloader) and/or submit a ProblemReport (send-pr). Regards, Adrian. > > I cannot upgrade it to 7.x, must use 6.x > > Regards, > Robert Heron > > _______________________________________________ > freebsd-smp@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-smp > To unsubscribe, send any mail to "freebsd-smp-unsubscribe@freebsd.org" > From won.derick at yahoo.com Tue Mar 10 07:55:49 2009 From: won.derick at yahoo.com (Won De Erick) Date: Tue Mar 10 07:55:55 2009 Subject: only one logical CPU used in Xeon Message-ID: <719357.64276.qm@web45802.mail.sp1.yahoo.com> Hi, I also have boxes with several Intel Xeon dual-core processors, and also got same results. With FreeBSD 6.2R: #sysctl -a | grep pci.enable hw.pci.enable_io_modes: 1 I think 6.2 uses legacy interrupt (INTx) which allows sharing of IRQs and is CPU bounded(?). It is also probable that even MSI/MSIX-supported NIC drivers are used here, 6.2R still used the traditional interrupt mechanism. With FreeBSD 7.1: #sysctl -a | grep pci.enable hw.pci.enable_msix: 1 hw.pci.enable_msi: 1 hw.pci.enable_io_modes: 1 Google would tell us that MSI/MSI-X allows control over which processor is chosen as the interrupt target. Does this explain the result? http://www.onlamp.com/pub/a/bsd/2008/02/26/whats-new-in-freebsd-70.html?page=last Regards, Won --- On Tue, 3/10/09, Robert Heron wrote: > Hi, > > I have FreeBSD 6.2R running on Intel SE7501HG2 motherboard > with two Xeon installed. > System detects 2 CPUs with 2 cores in each of them (total 4 > cores) but it uses only one core in each Xeon (total 2 of 4 > cores are used only). > I see it in the 'top'. I tried the same hardware on FreeBSD > 7.1R and this OS uses all of the detected cores. > > What should I do FreeBSD 6.2R to use all of the available > cores? > I cannot upgrade it to 7.x, must use 6.x > > Regards, > Robert Heron > From jhb at freebsd.org Tue Mar 10 08:14:39 2009 From: jhb at freebsd.org (John Baldwin) Date: Tue Mar 10 08:14:46 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <719357.64276.qm@web45802.mail.sp1.yahoo.com> References: <719357.64276.qm@web45802.mail.sp1.yahoo.com> Message-ID: <200903101114.20634.jhb@freebsd.org> On Tuesday 10 March 2009 10:43:01 am Won De Erick wrote: > > Hi, > > I also have boxes with several Intel Xeon dual-core processors, and also got same results. > > With FreeBSD 6.2R: > #sysctl -a | grep pci.enable > hw.pci.enable_io_modes: 1 > > I think 6.2 uses legacy interrupt (INTx) which allows sharing of IRQs and is CPU bounded(?). It is also probable that even MSI/MSIX-supported NIC drivers are used here, 6.2R still used the traditional interrupt mechanism. > > With FreeBSD 7.1: > #sysctl -a | grep pci.enable > hw.pci.enable_msix: 1 > hw.pci.enable_msi: 1 > hw.pci.enable_io_modes: 1 > > Google would tell us that MSI/MSI-X allows control over which processor is chosen as the interrupt target. Does this explain the result? MSI won't determine which CPUs are used. Probably the submitter needs to try 6.4. My guess is that 6.2 thinks his extra cores are hyperthreads and thus has them disabled by default. You can toggle machdep.hyperthreading_allowed (sysctl) if that is the case, but upgrading to 6.4 is probably better. -- John Baldwin From robert at heron.pl Tue Mar 10 09:28:21 2009 From: robert at heron.pl (Robert Heron) Date: Tue Mar 10 09:28:28 2009 Subject: only one logical CPU used in Xeon In-Reply-To: References: <2E7DD8D9-FABA-4A30-9118-948D26DB26AE@heron.pl> Message-ID: <8CFE3F10-2504-45DB-9ED7-BDF354BC10FF@heron.pl> Hyperthreading is enabled in BIOS, but the variable: machdep.hyperthreading_allowed is '0' by default in FreeBSD 6.2R :((( I just changed it now to '1' and all cores in all CPUs started to work :))) Very simple change and so much power more :) Thanks a lot for this hint!! Regards, Robert On Mar 10, 2009, at 3:19 PM, Lucas Holt wrote: > Have you tried the hyperthreading setting? > > machdep.hyperthreading_allowed=1 > (put that in /etc/sysctl.conf) > > On Mar 10, 2009, at 6:40 AM, Robert Heron wrote: > >> Hi, >> >> I have FreeBSD 6.2R running on Intel SE7501HG2 motherboard with two >> Xeon installed. >> System detects 2 CPUs with 2 cores in each of them (total 4 cores) >> but it uses only one core in each Xeon (total 2 of 4 cores are used >> only). >> I see it in the 'top'. I tried the same hardware on FreeBSD 7.1R >> and this OS uses all of the detected cores. >> >> What should I do FreeBSD 6.2R to use all of the available cores? >> I cannot upgrade it to 7.x, must use 6.x >> >> Regards, >> Robert Heron >> >> _______________________________________________ >> freebsd-smp@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-smp >> To unsubscribe, send any mail to "freebsd-smp- >> unsubscribe@freebsd.org" > > > > Lucas Holt > Luke@FoolishGames.com > ________________________________________________________ > MidnightBSD.org (Free OS) > JustJournal.com (Free blogging) > > From julian at elischer.org Tue Mar 10 09:40:25 2009 From: julian at elischer.org (Julian Elischer) Date: Tue Mar 10 09:40:31 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <8CFE3F10-2504-45DB-9ED7-BDF354BC10FF@heron.pl> References: <2E7DD8D9-FABA-4A30-9118-948D26DB26AE@heron.pl> <8CFE3F10-2504-45DB-9ED7-BDF354BC10FF@heron.pl> Message-ID: <49B697F8.3070301@elischer.org> Robert Heron wrote: > > Hyperthreading is enabled in BIOS, but the variable: > > machdep.hyperthreading_allowed > > is '0' by default in FreeBSD 6.2R :((( > I just changed it now to '1' and all cores in all CPUs started to work :))) > > Very simple change and so much power more :) > well certainly more power used.. It would be a good idea to do benchmarks on your workload to check if it actually helps. This is is because Hyperthreading on some cpu models actually reduces throughput on some workloads because it thrashes the shared cache more. Remember that a secondary cache miss can take up as much extra time as dozens of instructions. From olli at lurza.secnetix.de Tue Mar 10 14:31:39 2009 From: olli at lurza.secnetix.de (Oliver Fromme) Date: Tue Mar 10 14:31:46 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <49B697F8.3070301@elischer.org> Message-ID: <200903102131.n2ALVIjR072494@lurza.secnetix.de> Julian Elischer wrote: > Robert Heron wrote: > > Hyperthreading is enabled in BIOS, but the variable: > > > > machdep.hyperthreading_allowed > > > > is '0' by default in FreeBSD 6.2R :((( > > I just changed it now to '1' and all cores in all CPUs started to work :))) > > > > Very simple change and so much power more :) > > well certainly more power used.. That's true. > It would be a good idea to do benchmarks on your workload > to check if it actually helps. This is is because > Hyperthreading on some cpu models actually reduces throughput > on some workloads because it thrashes the shared cache more. No need for benchmarks here, because Robert's processors don't do hyperthreading at all. FreeBSD 6.2 recognized the *cores* as logical HT processors, so he needs to set hyperthreading_allowed in order to enable the cores. Of course, the better solution is to update to a more recent release, such as 6.4 or even 7.1. His 6.2 is more than 2 years old and contains lots of known bugs, of which the wrong recognition of cores is a relatively harmless one. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M. Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung: secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n- chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd "That's what I love about GUIs: They make simple tasks easier, and complex tasks impossible." -- John William Chambless From robert at heron.pl Wed Mar 11 00:36:52 2009 From: robert at heron.pl (Robert Heron) Date: Wed Mar 11 00:36:59 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <200903102131.n2ALVIjR072494@lurza.secnetix.de> References: <200903102131.n2ALVIjR072494@lurza.secnetix.de> Message-ID: On Mar 10, 2009, at 10:31 PM, Oliver Fromme wrote: > No need for benchmarks here, because Robert's processors > don't do hyperthreading at all. FreeBSD 6.2 recognized > the *cores* as logical HT processors, so he needs to > set hyperthreading_allowed in order to enable the cores. > > Of course, the better solution is to update to a more > recent release, such as 6.4 or even 7.1. His 6.2 is > more than 2 years old and contains lots of known bugs, > of which the wrong recognition of cores is a relatively > harmless one. I also tried 6.4 with exactly the same wrong results as 6.2. 7.1 is OK, but is has hyperthreading_allowed=1 by default. Robert From olli at lurza.secnetix.de Wed Mar 11 01:03:46 2009 From: olli at lurza.secnetix.de (Oliver Fromme) Date: Wed Mar 11 01:03:53 2009 Subject: only one logical CPU used in Xeon In-Reply-To: Message-ID: <200903110803.n2B83SxX097243@lurza.secnetix.de> Robert Heron wrote: > Oliver Fromme wrote: > > No need for benchmarks here, because Robert's processors > > don't do hyperthreading at all. FreeBSD 6.2 recognized > > the *cores* as logical HT processors, so he needs to > > set hyperthreading_allowed in order to enable the cores. > > > > Of course, the better solution is to update to a more > > recent release, such as 6.4 or even 7.1. His 6.2 is > > more than 2 years old and contains lots of known bugs, > > of which the wrong recognition of cores is a relatively > > harmless one. > > I also tried 6.4 with exactly the same wrong results as 6.2. That's suprising. Actually I think that all fixes to the CPU detection code have been backported to 6-stable. > 7.1 is OK, but is has hyperthreading_allowed=1 by default. What happens if you set hyperthreading_allowed=0 on 7.1? Are you 100% sure that your processor are really multi-core and not just hyperthreaded? Please post the relevant lines from /var/run/dmesg.boot or output from the `dmesg` command (as Adrian already suggested), preferably on a recent FreeBSD system (i.e. 7.1). The top 25 lines should be sufficient. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M. Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung: secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n- chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd "Documentation is like sex; when it's good, it's very, very good, and when it's bad, it's better than nothing." -- Dick Brandon From robert at heron.pl Wed Mar 11 03:21:53 2009 From: robert at heron.pl (Robert Heron) Date: Wed Mar 11 03:22:03 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <200903110803.n2B83SxX097243@lurza.secnetix.de> References: <200903110803.n2B83SxX097243@lurza.secnetix.de> Message-ID: <57F9AE2E-E5B2-4611-8094-B64C598D5DF4@heron.pl> On Mar 11, 2009, at 9:03 AM, Oliver Fromme wrote: >> I also tried 6.4 with exactly the same wrong results as 6.2. > > That's suprising. Actually I think that all fixes to the > CPU detection code have been backported to 6-stable. > Cores are detected correctly, but not used all by default (hyperthreading_allowed=0). When I change hyperthreading_allowed to 1 on a running system it starts to use all cores. When switch back to 0, uses only 1 core of each cpu. >> 7.1 is OK, but is has hyperthreading_allowed=1 by default. > > What happens if you set hyperthreading_allowed=0 on 7.1? > no change. 7.1 uses all cores regardless of value of hyperthreading_allowed > Are you 100% sure that your processor are really multi-core > and not just hyperthreaded? Please post the relevant lines > from /var/run/dmesg.boot or output from the `dmesg` command > (as Adrian already suggested), preferably on a recent > FreeBSD system (i.e. 7.1). The top 25 lines should be > sufficient. I have tried 6.2, 6.4 and 7.1 on two different servers. The first server of them (older) is Intel SE7501HG2 + 2 x Xeon 2.44GHz (2 cores in each Xeon) The second one (newer) is Intel S5000VSA + 2 x Xeon 2.66GHz (4 cores in each Xeon) ================= dmesg 6.2 on the newer hardware =================== > Copyright (c) 1992-2007 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 6.2-RELEASE #1: Sun Mar 8 20:02:38 CET 2009 root@[machine name removed]:/usr/src/sys/i386/compile/ HERON-62-2009-03-08 ACPI APIC Table: Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: Intel(R) Xeon(TM) CPU 2.66GHz (2660.02-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0xf64 Stepping = 4 Features = 0xbfebfbff < FPU ,VME ,DE ,PSE ,TSC ,MSR ,PAE ,MCE ,CX8 ,APIC ,SEP ,MTRR ,PGE ,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE> Features2=0xe4bd,> AMD Features=0x20100000 AMD Features2=0x1 Cores per package: 2 Logical CPUs per core: 2 real memory = 2680160256 (2556 MB) avail memory = 2622218240 (2500 MB) FreeBSD/SMP: Multiprocessor System Detected: 8 CPUs cpu0 (BSP): APIC ID: 0 cpu1 (AP): APIC ID: 1 cpu2 (AP): APIC ID: 2 cpu3 (AP): APIC ID: 3 cpu4 (AP): APIC ID: 4 cpu5 (AP): APIC ID: 5 cpu6 (AP): APIC ID: 6 cpu7 (AP): APIC ID: 7 ioapic0 irqs 0-23 on motherboard ioapic1 irqs 24-47 on motherboard lapic0: Forcing LINT1 to edge trigger kbd1 at kbdmux0 acpi0: on motherboard acpi0: Power Button (fixed) Timecounter "ACPI-fast" frequency 3579545 Hz quality 1000 acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 cpu0: on acpi0 acpi_throttle0: on cpu0 cpu1: on acpi0 acpi_throttle1: on cpu1 acpi_throttle1: failed to attach P_CNT device_attach: acpi_throttle1 attach returned 6 cpu2: on acpi0 acpi_throttle2: on cpu2 acpi_throttle2: failed to attach P_CNT device_attach: acpi_throttle2 attach returned 6 cpu3: on acpi0 acpi_throttle3: on cpu3 acpi_throttle3: failed to attach P_CNT device_attach: acpi_throttle3 attach returned 6 cpu4: on acpi0 acpi_throttle4: on cpu4 acpi_throttle4: failed to attach P_CNT device_attach: acpi_throttle4 attach returned 6 cpu5: on acpi0 acpi_throttle5: on cpu5 acpi_throttle5: failed to attach P_CNT device_attach: acpi_throttle5 attach returned 6 cpu6: on acpi0 acpi_throttle6: on cpu6 acpi_throttle6: failed to attach P_CNT device_attach: acpi_throttle6 attach returned 6 cpu7: on acpi0 acpi_throttle7: on cpu7 acpi_throttle7: failed to attach P_CNT device_attach: acpi_throttle7 attach returned 6 ================== dmesg 7.1 on the older hardware ==================== Copyright (c) 1992-2009 The FreeBSD Project. Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 The Regents of the University of California. All rights reserved. FreeBSD is a registered trademark of The FreeBSD Foundation. FreeBSD 7.1-RELEASE #1: Wed Mar 11 10:38:08 CET 2009 root@[machine name removed]:/usr/src/sys/i386/compile/HERON Timecounter "i8254" frequency 1193182 Hz quality 0 CPU: Intel(R) Xeon(TM) CPU 2.40GHz (2392.29-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0xf25 Stepping = 5 Features = 0xbfebfbff < FPU ,VME ,DE ,PSE ,TSC ,MSR ,PAE ,MCE ,CX8 ,APIC ,SEP ,MTRR ,PGE ,MCA,CMOV,PAT,PSE36,CLFLUSH,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE> Features2=0x4400 Logical CPUs per core: 2 real memory = 4026466304 (3839 MB) avail memory = 3942027264 (3759 MB) ACPI APIC Table: FreeBSD/SMP: Multiprocessor System Detected: 4 CPUs cpu0 (BSP): APIC ID: 0 cpu1 (AP): APIC ID: 1 cpu2 (AP): APIC ID: 6 cpu3 (AP): APIC ID: 7 ioapic0 irqs 0-23 on motherboard ioapic1 irqs 24-47 on motherboard ioapic2 irqs 48-71 on motherboard lapic0: Forcing LINT1 to edge trigger kbd1 at kbdmux0 acpi0: on motherboard ACPI Warning (dswload-0794): Type override - [DEB_] had invalid type (Integer) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [MLIB] had invalid type (Integer) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [DATA] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [SIO_] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [LEDP] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [GPEN] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [GPST] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [GP1N] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [WUES] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [WUSE] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [SBID] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [SWCE] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ACPI Warning (dswload-0794): Type override - [SMIR] had invalid type (String) for Scope operator, changed to (Scope) [20070320] ======================================================================== I observed that hardware does not make any difference. 6.2 and 6.4 uses all cores when machdep.hyperthreading_allowed=1 7.1 uses all cores regardless of value of machdep.hyperthreading_allowed variable. Regards, Robert From jhb at freebsd.org Wed Mar 11 06:55:35 2009 From: jhb at freebsd.org (John Baldwin) Date: Wed Mar 11 06:55:41 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <57F9AE2E-E5B2-4611-8094-B64C598D5DF4@heron.pl> References: <200903110803.n2B83SxX097243@lurza.secnetix.de> <57F9AE2E-E5B2-4611-8094-B64C598D5DF4@heron.pl> Message-ID: <200903110837.57483.jhb@freebsd.org> On Wednesday 11 March 2009 6:21:50 am Robert Heron wrote: > > Are you 100% sure that your processor are really multi-core > > and not just hyperthreaded? Please post the relevant lines > > from /var/run/dmesg.boot or output from the `dmesg` command > > (as Adrian already suggested), preferably on a recent > > FreeBSD system (i.e. 7.1). The top 25 lines should be > > sufficient. > > I have tried 6.2, 6.4 and 7.1 on two different servers. The first > server of them (older) is Intel SE7501HG2 + 2 x Xeon 2.44GHz (2 cores > in each Xeon) > The second one (newer) is Intel S5000VSA + 2 x Xeon 2.66GHz (4 cores > in each Xeon) The older box does have hyperthreading and the new one does not, so 6.x is working properly. The issue with 7.1 is that if you use the default scheduler (ULE) it doesn't actually honor the hyperthreading_allowed variable. 7.x after 7.1 has been changed to at least honor it as a tunable in the ULE case (though the sysctl doesn't work). If you use 4BSD on 7.1 then the tunable and sysctl will work fine. -- John Baldwin From olli at lurza.secnetix.de Wed Mar 11 08:02:08 2009 From: olli at lurza.secnetix.de (Oliver Fromme) Date: Wed Mar 11 08:02:16 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <57F9AE2E-E5B2-4611-8094-B64C598D5DF4@heron.pl> Message-ID: <200903111501.n2BF1pIj016949@lurza.secnetix.de> Robert Heron wrote: > I have tried 6.2, 6.4 and 7.1 on two different servers. The first > server of them (older) is Intel SE7501HG2 + 2 x Xeon 2.44GHz (2 cores > in each Xeon) > The second one (newer) is Intel S5000VSA + 2 x Xeon 2.66GHz (4 cores > in each Xeon) I'm afraid that's wrong. According to your dmesg output, the older machine has two Xeon processors which support hyperthreading, so you have a total of four logical CPUs (but only two physical). These Xeons are *not* multi-core, they're single-core. The newer machine has two Xeon processors with two cores each (not four!), again with hyperthreading, so you have a total of eight logical CPUs. So everything is working as expected. Here's how you can see the details in the demsg output: The older machine: > Features = 0xbfebfbff <...,HTT,...> The "HTT" bit in the features bitmask indicates that the processor supports hyperthreading. > Logical CPUs per core: 2 This also refers to hyperthreading: You have two logical (i.e. hyperthreaded) CPUs per core. > FreeBSD/SMP: Multiprocessor System Detected: 4 CPUs That's the total number of logical CPUs: You have two processor packages, one core per package, and two HTT- CPUs per core: 2 * 1 * 2 == 4. But you only have two physical cores (one per processor package). Now the newer machine: > Features = 0xbfebfbff <...,HTT,...> Again: "HTT" == hyperthreading supported. > Cores per package: 2 These are a multi-core processor: There are two cores per processor package. > Logical CPUs per core: 2 And again hyperthreading: Two logical CPUs per core. > FreeBSD/SMP: Multiprocessor System Detected: 8 CPUs So you get a total of eight logical CPUs, including hyperthreading (2 * 2 * 2 == 8). Note that there are only four physical cores. Whether hyperthreading will improve performance is controversial. Most people seem to believe that it depends on your kind of application, but one thing is sure: hyperthreading will *not* double the performance, not even nearly. In most cases it's a matter of a few percent only, probably barely noticeable. And in some cases hyperthreading will make things worse. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M. Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung: secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n- chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd "The scanf() function is a large and complex beast that often does something almost but not quite entirely unlike what you desired." -- Chris Torek From jhb at freebsd.org Thu Mar 12 06:58:45 2009 From: jhb at freebsd.org (John Baldwin) Date: Thu Mar 12 06:58:51 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <200903111501.n2BF1pIj016949@lurza.secnetix.de> References: <200903111501.n2BF1pIj016949@lurza.secnetix.de> Message-ID: <200903120844.13506.jhb@freebsd.org> On Wednesday 11 March 2009 11:01:51 am Oliver Fromme wrote: > Here's how you can see the details in the demsg output: > The older machine: > > > Features = 0xbfebfbff <...,HTT,...> > > The "HTT" bit in the features bitmask indicates that the > processor supports hyperthreading. No, that flag just means it supports a register where you can ask for the total number of threads/cores per package. Multi-core CPUs that don't have hyperthreads have 'HTT' set, but when you read the registers properly you see that they have 1 thread per core. -- John Baldwin From won.derick at yahoo.com Thu Mar 12 07:28:16 2009 From: won.derick at yahoo.com (won.derick@yahoo.com) Date: Thu Mar 12 07:28:22 2009 Subject: only one logical CPU used in Xeon Message-ID: <418159.62247.qm@web45804.mail.sp1.yahoo.com> --- On Thu, 3/12/09, John Baldwin wrote: > > -----Inline Attachment Follows----- > > On Wednesday 11 March 2009 11:01:51 > am Oliver Fromme wrote: > > Here's how you can see the details in the demsg > output: > > The older machine: > > > >? > Features = 0xbfebfbff <...,HTT,...> > > > > The "HTT" bit in the features bitmask indicates that > the > > processor supports hyperthreading. > > No, that flag just means it supports a register where you > can ask for the > total number of threads/cores per package.? Multi-core > CPUs that don't have > hyperthreads have 'HTT' set, but when you read the > registers properly you see > that they have 1 thread per core. > Does this mean that the # of logical CPUs per core is not enough to determine whether the machine supports hyperthreading or not? You've said earlier that the newer machine is not hyperthreaded, so I'm assuming that there is another part in the dmesg output that says so. more guidance pls. From jhb at freebsd.org Thu Mar 12 10:14:54 2009 From: jhb at freebsd.org (John Baldwin) Date: Thu Mar 12 10:15:01 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <418159.62247.qm@web45804.mail.sp1.yahoo.com> References: <418159.62247.qm@web45804.mail.sp1.yahoo.com> Message-ID: <200903121055.23408.jhb@freebsd.org> On Thursday 12 March 2009 10:28:15 am won.derick@yahoo.com wrote: > > --- On Thu, 3/12/09, John Baldwin wrote: > > > > -----Inline Attachment Follows----- > > > > On Wednesday 11 March 2009 11:01:51 > > am Oliver Fromme wrote: > > > Here's how you can see the details in the demsg > > output: > > > The older machine: > > > > > >? > Features = 0xbfebfbff <...,HTT,...> > > > > > > The "HTT" bit in the features bitmask indicates that > > the > > > processor supports hyperthreading. > > > > No, that flag just means it supports a register where you > > can ask for the > > total number of threads/cores per package.? Multi-core > > CPUs that don't have > > hyperthreads have 'HTT' set, but when you read the > > registers properly you see > > that they have 1 thread per core. > > > > Does this mean that the # of logical CPUs per core is not enough to determine whether the machine supports hyperthreading or not? You've said earlier that the newer machine is not hyperthreaded, so I'm assuming that there is another part in the dmesg output that says so. more guidance pls. Your new machine might be hyperthreaded. I'm not really sure as it's a P4 descendant and I'm not sure if Intel released some dual-core P4 Xeon that had 2 threads per core. The current Core processors do not implement multiple threads in their cores, but neither of your systems have those CPUs. -- John Baldwin From won.derick at yahoo.com Tue Mar 17 00:24:50 2009 From: won.derick at yahoo.com (won.derick@yahoo.com) Date: Tue Mar 17 00:24:56 2009 Subject: only one logical CPU used in Xeon Message-ID: <916655.40447.qm@web45808.mail.sp1.yahoo.com> --- On Wed, 3/11/09, John Baldwin wrote: > in the ULE case (though the sysctl doesn't work). If > you use 4BSD on 7.1 > then the tunable and sysctl will work fine. > I tried this, but I can't see the variable being displayed. #uname -a FreeBSD 7.1-RELEASE FreeBSD 7.1-RELEASE #1: Mon Mar 16 12:06:49 PHT 2009 ... #sysctl -a | grep kern.sched.name kern.sched.name: 4BSD #sysctl -a | grep machdep machdep.acpi_timer_freq: 3579545 machdep.enable_panic_key: 0 machdep.adjkerntz: -28800 machdep.wall_cmos_clock: 1 machdep.disable_rtc_set: 0 machdep.acpi_root: 1040336 machdep.disable_mtrrs: 0 machdep.cpu_idle_hlt: 1 machdep.hlt_cpus: 0 machdep.prot_fault_translation: 0 machdep.panic_on_nmi: 1 machdep.tsc_freq: 2992517568 machdep.i8254_freq: 1193182 machdep.conspeed: 9600 machdep.gdbspeed: 9600 machdep.conrclk: 1843200 machdep.hlt_logical_cpus: 0 machdep.logical_cpus_mask: 14 How should make the variable visible? I am using IBM x3650 with a Quad-core Xeon processor. Thank you very much. Won From olli at lurza.secnetix.de Tue Mar 17 05:31:18 2009 From: olli at lurza.secnetix.de (Oliver Fromme) Date: Tue Mar 17 05:31:26 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <916655.40447.qm@web45808.mail.sp1.yahoo.com> Message-ID: <200903171230.n2HCUqjB080921@lurza.secnetix.de> won.derick@yahoo.com wrote: > John Baldwin wrote: > > in the ULE case (though the sysctl doesn't work). If > > you use 4BSD on 7.1 > > then the tunable and sysctl will work fine. > > > > I tried this, but I can't see the variable being displayed. Then your machine doesn't support hyperthreading. The machdep.hyperthreading_allowed sysctl is only present if your machine actually has multiple logical CPUs. If you don't see that sysctl, then your processor doesn't have multiple hyperthreaded CPUs. If you're interested how this is done in the source, see the function cpu_hlt_setup() in src/sys/amd64/amd64/mp_machdep.c or src/sys/i386/i386/mp_machdep.c, respectively. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M. Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung: secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n- chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd C++: "an octopus made by nailing extra legs onto a dog" -- Steve Taylor, 1998 From won.derick at yahoo.com Tue Mar 17 07:14:03 2009 From: won.derick at yahoo.com (Won De Erick) Date: Tue Mar 17 07:14:10 2009 Subject: only one logical CPU used in Xeon Message-ID: <100058.54621.qm@web45803.mail.sp1.yahoo.com> --- On Tue, 3/17/09, Oliver Fromme wrote: > won.derick@yahoo.com wrote: > > John Baldwin wrote: > > > in the ULE case (though the sysctl doesn't work).? If > > > you use 4BSD on 7.1 > > > then the tunable and sysctl will work fine. > > > > > > > I tried this, but I can't see the variable being > > displayed. > > Then your machine doesn't support hyperthreading. > > The machdep.hyperthreading_allowed sysctl is only present > if your machine actually has multiple logical CPUs. > If you don't see that sysctl, then your processor doesn't > have multiple hyperthreaded CPUs. > Thanks for that further clarification. The dmesg has also justified this. #dmesg ... Cores per package: 4 ... FreeBSD/SMP: Multiprocessor System Detected: 4 CPUs ... > If you're interested how this is done in the source, see > the > function cpu_hlt_setup() in > src/sys/amd64/amd64/mp_machdep.c > or src/sys/i386/i386/mp_machdep.c, respectively. > From olli at lurza.secnetix.de Tue Mar 17 13:51:39 2009 From: olli at lurza.secnetix.de (Oliver Fromme) Date: Tue Mar 17 13:51:46 2009 Subject: only one logical CPU used in Xeon In-Reply-To: <100058.54621.qm@web45803.mail.sp1.yahoo.com> Message-ID: <200903172051.n2HKpEiN005869@lurza.secnetix.de> Won De Erick wrote: > Thanks for that further clarification. The dmesg has also justified this. > > #dmesg > ... > Cores per package: 4 > ... > FreeBSD/SMP: Multiprocessor System Detected: 4 CPUs > ... Yes. On machines with hyperthreaded processors, you would also see the message "Logical CPUs per core: 2" or similar in dmesg. This message appears only if you have hyper- threaded processors. Best regards Oliver -- Oliver Fromme, secnetix GmbH & Co. KG, Marktplatz 29, 85567 Grafing b. M. Handelsregister: Registergericht Muenchen, HRA 74606, Gesch?ftsfuehrung: secnetix Verwaltungsgesellsch. mbH, Handelsregister: Registergericht M?n- chen, HRB 125758, Gesch?ftsf?hrer: Maik Bachmann, Olaf Erb, Ralf Gebhart FreeBSD-Dienstleistungen, -Produkte und mehr: http://www.secnetix.de/bsd "File names are infinite in length, where infinity is set to 255 characters." -- Peter Collinson, "The Unix File System"