Re: numa and nodes

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, Apr 30, 2007 at 03:39:34AM +0200, ninjaboy wrote:
> I'm reading about NUMA, and i've read numa_maps in proc of some process,
> why the kernel uses only one node?

- From a performance point of view you want that because memory access
from other nodes is more expensive than the memory access from the
local node.

> # cat /proc/self/numa_maps
> 00400000 default file=/bin/cat mapped=3 N0=3
> 
> every process has always N0, that mean node 0
> I'm on amd turion64 X2, then I've dual core.

In case of an AMD64 CPU, dual core means that the physical CPU has two
processor cores which share the same memory controller and hence are on
the same node. Therefore all your processes will be in node 0.

Here's the output of "cat /proc/self/numa_maps" from a dual amd64 box
(i.e.: two single core CPUs):

  08048000 default file=/bin/cat mapped=4 N0=4
  0804c000 default file=/bin/cat anon=1 dirty=1 N0=1
  0804d000 default heap anon=1 dirty=1 active=0 N0=1
  f7e35000 default anon=1 dirty=1 N0=1
  f7e36000 default file=/lib/tls/i686/cmov/libc-2.3.2.so mapped=53 mapmax=58 N0=2 N1=51
  f7f60000 default file=/lib/tls/i686/cmov/libc-2.3.2.so anon=5 dirty=5 N0=5
  f7f69000 default anon=2 dirty=2 active=0 N0=2
  f7f74000 default anon=1 dirty=1 N0=1
  f7f75000 default file=/lib/ld-2.3.2.so mapped=20 mapmax=58 N1=20
  f7f8b000 default file=/lib/ld-2.3.2.so anon=1 dirty=1 N0=1
  fff00000 default stack anon=3 dirty=3 N0=3
  ffffe000 default mapped=1 mapmax=58 active=0 N0=1

> The unique node is not used only in x86 arch?

Sorry, no idea what you mean.


Erik

- -- 
They're all fools. Don't worry. Darwin may be slow, but he'll
eventually get them. -- Matthew Lammers in alt.sysadmin.recovery
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD4DBQFGN2zZ/PlVHJtIto0RAn1IAJEBdBur4SCJBVjFwnYc+JgtMaR0AJ4hR+kQ
PXJDGWDtJ2GQnOUf+joXPQ==
=xA2s
-----END PGP SIGNATURE-----

--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx
Please read the FAQ at http://kernelnewbies.org/FAQ


[Index of Archives]     [Newbies FAQ]     [Linux Kernel Mentors]     [Linux Kernel Development]     [IETF Annouce]     [Git]     [Networking]     [Security]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux RAID]     [Linux SCSI]     [Linux ACPI]
  Powered by Linux