* WARNING: MP table in the EBDA can be UNSAFE
@ 2003-12-05 15:18 Paul Rolland
2003-12-05 16:00 ` Zwane Mwaikambo
0 siblings, 1 reply; 7+ messages in thread
From: Paul Rolland @ 2003-12-05 15:18 UTC (permalink / raw)
To: linux-smp, linux-kernel; +Cc: 'Paul Rolland'
Hello,
Our Linux, running on an IBM X-Series 445, says :
(excerpt from dmesg) :
Linux version 2.4.20 (root@stg-02.ntr.witbe.francetelecom.fr) (gcc version 2.96 20000731 (Red Hat Linux 7.1 2.96-98)) #12 SMP
Fri Dec 5 11:08:40 GMT 2003
BIOS-provided physical RAM map:
BIOS-e820: 0000000000000000 - 000000000009dc00 (usable)
BIOS-e820: 000000000009dc00 - 00000000000a0000 (reserved)
BIOS-e820: 00000000000e0000 - 0000000000100000 (reserved)
BIOS-e820: 0000000000100000 - 000000007ffb7000 (usable)
BIOS-e820: 000000007ffb7000 - 000000007ffbf800 (ACPI data)
BIOS-e820: 000000007ffbf800 - 0000000080000000 (reserved)
BIOS-e820: 00000000fec00000 - 0000000100000000 (reserved)
1151MB HIGHMEM available.
896MB LOWMEM available.
found SMP MP-table at 0009dd40
hm, page 0009d000 reserved twice.
hm, page 0009e000 reserved twice.
hm, page 0009e000 reserved twice.
hm, page 0009f000 reserved twice.
WARNING: MP table in the EBDA can be UNSAFE, contact linux-smp@vger.kernel.org if you experience SMP problems!
On node 0 totalpages: 524215
zone(0): 4096 pages.
zone(1): 225280 pages.
zone(2): 294839 pages.
ACPI: Searched entire block, no RSDP was found.
ACPI: RSDP located at physical address c00fdba0
RSD PTR v0 [IBM ]
__va_range(0x7ffbf780, 0x68): idx=8 mapped at ffff6000
ACPI table found: RSDT v1 [IBM SERVIGIL 0.4096]
__va_range(0x7ffbf700, 0x24): idx=8 mapped at ffff6000
__va_range(0x7ffbf700, 0x74): idx=8 mapped at ffff6000
ACPI table found: FACP v1 [IBM SERVIGIL 0.4096]
__va_range(0x7ffbf640, 0x24): idx=8 mapped at ffff6000
__va_range(0x7ffbf640, 0x9a): idx=8 mapped at ffff6000
ACPI table found: APIC v1 [IBM SERVIGIL 0.4096]
__va_range(0x7ffbf640, 0x9a): idx=8 mapped at ffff6000
LAPIC (acpi_id[0x0000] id[0x0] enabled[1])
CPU 0 (0x0000) enabledProcessor #0 Pentium 4(tm) XEON(tm) APIC version 16
LAPIC (acpi_id[0x0001] id[0x12] enabled[1])
CPU 1 (0x1200) enabled<4>Processor #18 INVALID - (Max ID: 16).
LAPIC (acpi_id[0x0002] id[0x20] enabled[1])
CPU 1 (0x2000) enabled<4>Processor #32 INVALID - (Max ID: 16).
LAPIC (acpi_id[0x0003] id[0x32] enabled[1])
CPU 1 (0x3200) enabled<4>Processor #50 INVALID - (Max ID: 16).
Is there any known solution to re-enable the CPU 1, CPU 2 and CPU 3 ?
Is this an IBM bug ?
Please reply privately, I'm not subscribed to linux-smp.
Thanks in advance,
Paul
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Blessed are they who have nothing to say and cannot be persuaded to say it.
-James Russell Lowell
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Paul Rolland, rol@witbe.net
Witbe.net SA
Directeur Associe
--
Please no HTML, I'm not a browser - Pas d'HTML, je ne suis pas un navigateur
"Some people dream of success... while others wake up and work hard at it"
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-05 15:18 WARNING: MP table in the EBDA can be UNSAFE Paul Rolland
@ 2003-12-05 16:00 ` Zwane Mwaikambo
2003-12-05 16:10 ` Paul Rolland
0 siblings, 1 reply; 7+ messages in thread
From: Zwane Mwaikambo @ 2003-12-05 16:00 UTC (permalink / raw)
To: Paul Rolland; +Cc: linux-smp, linux-kernel
On Fri, 5 Dec 2003, Paul Rolland wrote:
> Our Linux, running on an IBM X-Series 445, says :
> (excerpt from dmesg) :
>
> found SMP MP-table at 0009dd40
> hm, page 0009d000 reserved twice.
> hm, page 0009e000 reserved twice.
> hm, page 0009e000 reserved twice.
> hm, page 0009f000 reserved twice.
> WARNING: MP table in the EBDA can be UNSAFE, contact linux-smp@vger.kernel.org if you experience SMP problems!
This bit is ok and can be safely ignored.
> LAPIC (acpi_id[0x0000] id[0x0] enabled[1])
> CPU 0 (0x0000) enabledProcessor #0 Pentium 4(tm) XEON(tm) APIC version 16
>
> LAPIC (acpi_id[0x0001] id[0x12] enabled[1])
> CPU 1 (0x1200) enabled<4>Processor #18 INVALID - (Max ID: 16).
> LAPIC (acpi_id[0x0002] id[0x20] enabled[1])
> CPU 1 (0x2000) enabled<4>Processor #32 INVALID - (Max ID: 16).
> LAPIC (acpi_id[0x0003] id[0x32] enabled[1])
> CPU 1 (0x3200) enabled<4>Processor #50 INVALID - (Max ID: 16).
>
> Is there any known solution to re-enable the CPU 1, CPU 2 and CPU 3 ?
>
> Is this an IBM bug ?
Did you compile your kernel with the following option?
IBM x440 Summit/EXA support
CONFIG_X86_SUMMIT
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-05 16:00 ` Zwane Mwaikambo
@ 2003-12-05 16:10 ` Paul Rolland
2003-12-05 16:16 ` Zwane Mwaikambo
0 siblings, 1 reply; 7+ messages in thread
From: Paul Rolland @ 2003-12-05 16:10 UTC (permalink / raw)
To: 'Zwane Mwaikambo'; +Cc: linux-smp, linux-kernel
Hello Zwane,
> > WARNING: MP table in the EBDA can be UNSAFE, contact
> linux-smp@vger.kernel.org if you experience SMP problems!
>
> This bit is ok and can be safely ignored.
OK, so let's forget about this one.
> Did you compile your kernel with the following option?
> IBM x440 Summit/EXA support
>
> CONFIG_X86_SUMMIT
>
I can't find this option ? Is this part of the 2.4.x branch ?
Paul
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-05 16:10 ` Paul Rolland
@ 2003-12-05 16:16 ` Zwane Mwaikambo
2003-12-05 16:59 ` Paul Rolland
2003-12-08 10:23 ` Paul Rolland
0 siblings, 2 replies; 7+ messages in thread
From: Zwane Mwaikambo @ 2003-12-05 16:16 UTC (permalink / raw)
To: Paul Rolland; +Cc: linux-smp, linux-kernel
On Fri, 5 Dec 2003, Paul Rolland wrote:
> > Did you compile your kernel with the following option?
> > IBM x440 Summit/EXA support
> >
> > CONFIG_X86_SUMMIT
> >
> I can't find this option ? Is this part of the 2.4.x branch ?
Indeed it is, you need to turn on "Multi-node NUMA system support"
CONFIG_X86_NUMA
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-05 16:16 ` Zwane Mwaikambo
@ 2003-12-05 16:59 ` Paul Rolland
2003-12-08 10:23 ` Paul Rolland
1 sibling, 0 replies; 7+ messages in thread
From: Paul Rolland @ 2003-12-05 16:59 UTC (permalink / raw)
To: 'Zwane Mwaikambo', 'Paul Rolland'; +Cc: linux-smp, linux-kernel
Hello,
> Indeed it is, you need to turn on "Multi-node NUMA system support"
>
> CONFIG_X86_NUMA
>
Gee... I was in a 2.4.20 kernel, and the SUMMIT option is not present,
but it is in 2.4.23...
I'll have to go to the server and do the re-installation using this
solution. I'll tell you Monday if it's working this way.
Regards, and many thanks for the tip,
Paul
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-05 16:16 ` Zwane Mwaikambo
2003-12-05 16:59 ` Paul Rolland
@ 2003-12-08 10:23 ` Paul Rolland
2003-12-08 14:15 ` Zwane Mwaikambo
1 sibling, 1 reply; 7+ messages in thread
From: Paul Rolland @ 2003-12-08 10:23 UTC (permalink / raw)
To: 'Zwane Mwaikambo', 'Paul Rolland'; +Cc: linux-smp, linux-kernel
Hello,
> > > Did you compile your kernel with the following option?
> > > IBM x440 Summit/EXA support
> > >
> > > CONFIG_X86_SUMMIT
> > >
> > I can't find this option ? Is this part of the 2.4.x branch ?
>
> Indeed it is, you need to turn on "Multi-node NUMA system support"
>
> CONFIG_X86_NUMA
That was it !
Great, the machine now sees all its CPUs.
The BogoMips number is rather strange (200 for a Xeon 2.4 GHz), but
the machine is really fine now....
Regards,
Paul
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: WARNING: MP table in the EBDA can be UNSAFE
2003-12-08 10:23 ` Paul Rolland
@ 2003-12-08 14:15 ` Zwane Mwaikambo
0 siblings, 0 replies; 7+ messages in thread
From: Zwane Mwaikambo @ 2003-12-08 14:15 UTC (permalink / raw)
To: Paul Rolland; +Cc: 'Paul Rolland', linux-smp, linux-kernel
On Mon, 8 Dec 2003, Paul Rolland wrote:
> > Indeed it is, you need to turn on "Multi-node NUMA system support"
> >
> > CONFIG_X86_NUMA
>
> That was it !
>
> Great, the machine now sees all its CPUs.
>
> The BogoMips number is rather strange (200 for a Xeon 2.4 GHz), but
> the machine is really fine now....
I wonder if that's an artifact of the timesource, for my edification could
you send your new bootlogs?
Ta
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2003-12-08 14:16 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-12-05 15:18 WARNING: MP table in the EBDA can be UNSAFE Paul Rolland
2003-12-05 16:00 ` Zwane Mwaikambo
2003-12-05 16:10 ` Paul Rolland
2003-12-05 16:16 ` Zwane Mwaikambo
2003-12-05 16:59 ` Paul Rolland
2003-12-08 10:23 ` Paul Rolland
2003-12-08 14:15 ` Zwane Mwaikambo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).