-----Original Message----- From: Greg KH <gregkh@suse.de> To: Кузьминский Михаил <mikky_m@mail.ru> Date: Sun, 16 Aug 2009 09:50:05 -0700 Subject: Re: [opensuse-kernel] Nehalem kernel NUMA bug
On Sun, Aug 16, 2009 at 08:39:59PM +0400, Кузьминский Михаил wrote:
I've dual E5520 server w/Supermicro X8DTI mobo (last June 2009 1.0c bios version). Default OpenSuSE 11.1 x86-64 kernel (2.6.27.7-9) gives an error in /sys/devices/system/node directory - there is node0 and node2 subdirectories instead of node0 and node1 (smt is turned off). As a result numactl tools don't work. I don't find the message about this error in SuSE bugzilla "database".
This Nehalem/NUMA error is also known (AFAIK) for 2.6.29-6 default OpenSuSE kernel, and also in FC11 kernel 2.6.29, and in some CentOS 2.6 kernel; but there are kernels w/o this error :-)
What are the kernels without this error?
AFAIK Centos 5.3 w/2.6.18-128.2.1 is OK. 2.6.18-128.1.10 don't work. But I didn't check it myself ! May be default OpenSuSE 10.3 kernel also is OK (see opensuse--hardware--64 bit forum on suse site, thread w/Nehalem content in "thread name").
May be extraction from dmesg on my SuSE11.1/2.6.27.7-9 will be helpful: it show, that it was NODE1, but then node2 is appear (see below) ! (see also http://marc.info/?l=linux-netdev&m=124967917523109&w=2)
ACPI: SRAT BF79A4B0, 0150 (r1 041409 OEMSRAT 1 INTL 1) ACPI: SSDT BF79FAC0, 249F (r1 DpgPmm CpuPm 12 INTL 20051117) ACPI: Local APIC address 0xfee00000 SRAT: PXM 0 -> APIC 0 -> Node 0 SRAT: PXM 0 -> APIC 2 -> Node 0 SRAT: PXM 0 -> APIC 4 -> Node 0 SRAT: PXM 0 -> APIC 6 -> Node 0 SRAT: PXM 1 -> APIC 16 -> Node 1 SRAT: PXM 1 -> APIC 18 -> Node 1 SRAT: PXM 1 -> APIC 20 -> Node 1 SRAT: PXM 1 -> APIC 22 -> Node 1 SRAT: Node 0 PXM 0 0-a0000 SRAT: Node 0 PXM 0 100000-c0000000 SRAT: Node 0 PXM 0 100000000-1c0000000 SRAT: Node 2 PXM 257 1c0000000-340000000 (here !!) This is a BIOS bug which gets workarounded in latest SLE11 kotd branch with
On Sunday 16 August 2009 11:27:18 pm Кузьминский Михаил wrote: these patches: Tue May 19 19:05:08 MSD 2009 - astarikovskiy@suse.de - patches-arch/acpi_srat-pxm-rev-ia64.patch patches.arch/acpi_srat-pxm-rev-x86-64.patch patches.arch/acpi_srat-pxm-rev-store.patch: Fix SRAT v.1 incompatibility on Bull (bnc#503038). Not sure what the mainline status about this is. You may want to have a look at the patches and complain to your BIOS/machine vendor. Thomas -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-kernel+help@opensuse.org