Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

Subject: Re: [OMPI devel] OpenMPI, PLPA and Linux cpuset/cgroup support
From: Chris Samuel (csamuel_at_[hidden])
Date: 2009-07-24 04:14:25


Hi Bert,

----- "Bert Wesarg" <bert.wesarg_at_[hidden]> wrote:

> The Cpus_allowed* fields in /proc/<pid>/status are the same as
> sched_getaffinity returns and the /proc/<pid>/cpuset needs to be
> resolved, i.e. where is the cpuset fs mounted?

The convention is to mount it on /dev/cpuset.

Unfortunately you cannot mount both the cpuset and
the cgroup virtual filesystems concurrently and with
the more advanced features of control groups such as
device whitelisting, memory usage limits, etc (and the
fact that cpusets is now a subset of cgroups) you will
need to be able to cope with either of these being in use.

Certainly it's something I'm concerned with as I'm
thinking about refactoring the Torque cpuset support
to be more generic.

cheers,
Chris

-- 
Christopher Samuel - (03) 9925 4751 - Systems Manager
 The Victorian Partnership for Advanced Computing
 P.O. Box 201, Carlton South, VIC 3053, Australia
VPAC is a not-for-profit Registered Research Agency