Date: Wed, 05 Feb 2003 14:13:08 -0500 (EST) From: John Baldwin <jhb@FreeBSD.org> To: Brendon and Wendy <brendy33@attbi.com> Cc: freebsd-smp@freebsd.org Subject: Re: mp_machdep.c patch Message-ID: <XFMail.20030205141308.jhb@FreeBSD.org> In-Reply-To: <200302041759.23937.brendy33@attbi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05-Feb-2003 Brendon and Wendy wrote: > Yep, the apic ids are the same with/without the htt patch. > > When you say "is a problem with your motherboard" do you mean a physical > design flaw, a fault or a BIOS issue? It does run HTT under linux and winXP. > I dont have either of those OSes on the machine right now, so I cant dump the > diagnostic info. Linux did need the ACPI SMP information to find the > extra "cores". Ok, then eventually 5.x will support HTT on this machine. According to what I've read in the IA-32 docs, the APIC ID's of logical cores are supposed to be sequential, so your motherboard does seem at least somewhat buggy. Hrmm. > Thanks for the reply. > Brendon > > > On Tuesday 04 February 2003 11:28 am, you wrote: >> On 04-Feb-2003 Brendon and Wendy wrote: >> > Hi, >> > >> > Saw the announce for the mp_machdep.c patch for hyperthreading. >> > Installed, built kernel and world. Everything seems fine. >> > >> > Question - how does one determine whether the hyperthreading working? My >> > DMESG is below, as well as output from top. Booting with HTT disabled >> > looks the same to me. >> > >> > This is on a supermicro p4dce+ (i860) with 2x HTT 2.0G xeons. >> > >> > Cheers, >> > Brendon >> > >> > Copyright (c) 1992-2003 The FreeBSD Project. >> > Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994 >> > The Regents of the University of California. All rightands >> > reserved. FreeBSD 4.7-STABLE #2: Mon Feb 3 18:35:10 PST 2003 >> > root@bigboot.humphrey.world:/usr/obj/usr/src/sys/BRENDY-STABLE >> > Timecounter "i8254" frequency 1193182 Hz >> > CPU: Intel(R) XEON(TM) CPU 2.00GHz (1982.52-MHz 686-class CPU) >> > Origin = "GenuineIntel" Id = 0xf24 Stepping = 4 >> > >> > Features=0x3febfbff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE, >> >MCA,CMOV,PAT,PSE36,CLFLUS H,DTS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM> >> > Hyperthreading: 2 logical CPUs >> > real memory = 536805376 (524224K bytes) >> > avail memory = 517742592 (505608K bytes) >> > Programming 24 pins in IOAPIC #0 >> > IOAPIC #0 intpin 2 -> irq 0 >> > FreeBSD/SMP: Multiprocessor motherboard >> > cpu0 (BSP): apic id: 0, version: 0x00050014, at 0xfee00000 >> > cpu1 (AP): apic id: 1, version: 0x00050014, at 0xfee00000 >> > io0 (APIC): apic id: 2, version: 0x00178020, at 0xfec00000 >> >> This set of APIC ID's is very odd and is your problem. :( Is your >> AP on apic id 1 when you booted up prior to having the HTT support? >> My guess is that it is. According to Intel, when you have HTT cores, >> the APIC ID's of the cores in a physical processor are sequential. >> The typical SMP + HTT setup I've seen is a physical cpu at ID 0 >> with it's 2nd core at ID 1 and the 2nd physical cpu at ID 6 with >> it's 2nd core at ID 7. The HTT core enumeration code assumes that >> this is the case, and if it finds that any of the APIC ID's that >> should be HTT cores are already in use by another physical CPU, it >> doesn't turn on any of the HTT cores because it has no way of >> talking to them. :) >> >> My guess is that this is a problem with your motherboard. You still >> have SMP, but it is just with one core per processor. > -- John Baldwin <jhb@FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/ "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.20030205141308.jhb>