SMP only detects one CPU - help?
Adam J Richardson
fatman.uk at gmail.com
Fri Apr 20 16:42:02 UTC 2007
Firstly, apologies for the length of this. It's my first mailing list
posting, so I hope I'm not breaching any etiquette codes. Also I've only
been using FreeBSD for a couple of years, so I'm still very new to the
I have recently become the proud owner of a second-hand HP NetServer
e800, which has two Pentium-III class  CPUs in it. I've installed
6.2-RELEASE-p3 [which is also referred to as 6.2-STABLE, is that
correct?] and built myself a SMP kernel. I'm sure it's a SMP kernel,
because the config file has "options SMP" in it and while running it
"sysctl kern.smp.maxcpus" returns 16. The BIOS says there are two CPUs,
so I guess the second CPU is in fine condition and ready to go.
I've done a little Googling and digging around in the system. First, the
obligatory "uname -a" output:
"FreeBSD boring.dnsalias.com 6.2-RELEASE-p3 FreeBSD 6.2-RELEASE-p3 #0:
Tue Apr 17 15:01:06 BST 2007
fatman at boring.dnsalias.com:/usr/obj/usr/src/sys/GAKKO i386"
There's nothing in /var/log/messages about CPUs, as far as I can tell.
The output of "dmesg" provides a possible clue:
"real memory = 268369920 (255 MB)
avail memory = 257167360 (245 MB)
MADT: Forcing active-low polarity and level trigger for SCI
ioapic0 <Version 1.1> irqs 0-15 on motherboard
ioapic1 <Version 1.1> irqs 16-31 on motherboard
kbd1 at kbdmux0
acpi0: <HP HWPC20F> on motherboard
acpi0: Power Button (fixed)
unknown: I/O range not supported
can't fetch resources for \\_SB_.PCI0.ISA_.SIO_.LPT_ -
Timecounter "ACPI-safe" frequency 3579545 Hz quality 1000
acpi_timer0: <32-bit timer at 3.579545MHz> port 0x1208-0x120b on acpi0
cpu0: <ACPI CPU> on acpi0
pcib0: <ACPI Host-PCI bridge> port 0xcf8-0xcff on acpi0
pci_link0: BIOS IRQ 10 for 0.15.INTA is invalid"
Several possible clues in that snippet, now that I look at it. The
"MADT" line is suspicious but I think the real culprit might be revealed
in the "can't fetch resources" line. On the other hand it might be a
completely unrelated problem. Google is silent on the subject of
"AE_AML_INVALID_RESOURCE_TYPE", and I don't see it in the Lehey or Lucas
Finally, mptable shows what's really going on:
"MP Config Base Table Entries:
Processors: APIC ID Version State Family Model Step Flags
3 0x11 BSP, usable 6 8 3 0x387fbff
0 0x11 AP, unusable 6 8 3 0x387fbff"
Apparently the "application processor" is "unusable", whatever that
means. I am now stuck.
Is there a software fix for this, or do I have a broken CPU, unlikely as
it seems? Any help is appreciated.
Does it seem likely that upgrading both CPUs to the highest Pentium-IIIs
I can find, which I was thinking about doing anyway, would fix the problem?
Adam J Richardson
More information about the freebsd-questions