				CPUSETS
				-------

Copyright (C) 2004 BULL SA.
Written by Simon.Derr@bull.net

Portions Copyright (c) 2004-2006 Silicon Graphics, Inc.
Modified by Paul Jackson <pj@sgi.com>
Modified by Christoph Lameter <clameter@sgi.com>

CONTENTS:
=========

1. Cpusets
  1.1 What are cpusets ?
  1.2 Why are cpusets needed ?
  1.3 How are cpusets implemented ?
  1.4 What are exclusive cpusets ?
  1.5 What does notify_on_release do ?
  1.6 What is memory_pressure ?
  1.7 What is memory spread ?
  1.8 How do I use cpusets ?
2. Usage Examples and Syntax
  2.1 Basic Usage
  2.2 Adding/removing cpus
  2.3 Setting flags
  2.4 Attaching processes
3. Questions
4. Contact

1. Cpusets
==========

1.1 What are cpusets ?
----------------------

Cpusets provide a mechanism for assigning a set of CPUs and Memory
Nodes to a set of tasks.   In this document "Memory Node" refers to
an on-line node that contains memory.

Cpusets constrain the CPU and Memory placement of tasks to only
the resources within a tasks current cpuset.  They form a nested
hierarchy visible in a virtual file system.  These are the essential
hooks, beyond what is already present, required to manage dynamic
job placement on large systems.

Each task has a pointer to a cpuset.  Multiple tasks may reference
the same cpuset.  Requests by a task, using the sched_setaffinity(2)
system call to include CPUs in its CPU affinity mask, and using the
mbind(2) and set_mempolicy(2) system calls to include Memory Nodes
in its memory policy, are both filtered through that tasks cpuset,
filtering out any CPUs or Memory Nodes not in that cpuset.  The
scheduler will not schedule a task on a CPU that is not allowed in
its cpus_allowed vector, and the kernel page allocator will not
allocate a page on a node that is not allowed in the requesting tasks
mems_allowed vector.

User level code may create and destroy cpusets by name in the cpuset
virtual file system, manage the attributes and permissions of these
cpusets and which CPUs and Memory Nodes are assigned to each cpuset,
specify and query to which cpuset a task is assigned, and list the
task pids assigned to a cpuset.


1.2 Why are cpusets needed ?
----------------------------

The management of large computer systems, with many processors (CPUs),
complex memory cache hierarchies and multiple Memory Nodes having
non-uniform access times (NUMA) presents additional challenges for
the efficient scheduling and memory placement of processes.

Frequently more modest sized systems can be operated with adequate
efficiency just by letting the operating system automatically share
the available CPU and Memory resources amongst the requesting tasks.

But larger systems, which benefit more from careful processor and
memory placement to reduce memory access times and contention,
and which typically represent a larger investment for the customer,
can benefit from explicitly placing jobs on properly sized subsets of
the system.

This can be especially valuable on:

    * Web Servers running multiple instances of the same web application,
    * Servers running different applications (for instance, a web server
      and a database), or
    * NUMA systems running large HPC applications with demanding
      performance characteristics.
    * Also cpu_exclusive cpusets are useful for servers running orthogonal
      workloads such as RT applications requiring low latency and HPC
      applications that are throughput sensitive

These subsets, or "soft partitions" must be able to be dynamically
adjusted, as the job mix changes, without impacting other concurrently
executing jobs. The location of the running jobs pages may also be moved
when the memory locations are changed.

The kernel cpuset patch provides the minimum essential kernel
mechanisms required to efficiently implement such subsets.  It
leverages existing CPU and Memory Placement facilities in the Linux
kernel to avoid any additional impact on the critical scheduler or
memory allocator code.


1.3 How are cpusets implemented ?
---------------------------------

Cpusets provide a Linux kernel mechanism to constrain which CPUs and
Memory Nodes are used by a process or set of processes.

The Linux kernel already has a pair of mechanisms to specify on which
CPUs a task may be scheduled (sched_setaffinity) and on which Memory
Nodes it may obtain memory (mbind, set_mempolicy).

Cpusets extends these two mechanisms as follows:

 - Cpusets are sets of allowed CPUs and Memory Nodes, known to the
   kernel.
 - Each task in the system is attached to a cpuset, via a pointer
   in the task structure to a reference counted cpuset structure.
 - Calls to sched_setaffinity are filtered to just those CPUs
   allowed in that tasks cpuset.
 - Calls to mbind and set_mempolicy are filtered to just
   those Memory Nodes allowed in that tasks cpuset.
 - The root cpuset contains all the systems CPUs and Memory
   Nodes.
 - For any cpuset, one can define child cpusets containing a subset
   of the parents CPU and Memory Node resources.
 - The hierarchy of cpusets can be mounted at /dev/cpuset, for
   browsing and manipulation from user space.
 - A cpuset may be marked exclusive, which ensures that no other
   cpuset (except direct ancestors and descendents) may contain
   any overlapping CPUs or Memory Nodes.
   Also a cpu_exclusive cpuset would be associated with a sched
   domain.
 - You can list all the tasks (by pid) attached to any cpuset.

The implementation of cpusets requires a few, simple hooks
into the rest of the kernel, none in performance critical paths:

 - in init/main.c, to initialize the root cpuset at system boot.
 - in fork and exit, to attach and detach a task from its cpuset.
 - in sched_setaffinity, to mask the requested CPUs by what's
   allowed in that tasks cpuset.
 - in sched.c migrate_all_tasks(), to keep migrating tasks within
   the CPUs allowed by their cpuset, if possible.
 - in sched.c, a new API partition_sched_domains for handling
   sched domain changes associated with cpu_exclusive cpusets
   and related changes in both sched.c and arch/ia64/kernel/domain.c
 - in the mbind and set_mempolicy system calls, to mask the requested
   Memory Nodes by what's allowed in that tasks cpuset.
 - in page_alloc.c, to restrict memory to allowed nodes.
 - in vmscan.c, to restrict page recovery to the current cpuset.

In addition a new file system, of type "cpuset" may be mounted,
typically at /dev/cpuset, to enable browsing and modifying the cpusets
presently known to the kernel.  No new system calls are added for
cpusets - all support for querying and modifying cpusets is via
this cpuset file system.

Each task under /proc has an added file named 'cpuset', displaying
the cpuset name, as the path relative to the root of the cpuset file
system.

The /proc/<pid>/status file for each task has two added lines,
displaying the tasks cpus_allowed (on which CPUs it may be scheduled)
and mems_allowed (on which Memory Nodes it may obtain memory),
in the format seen in the following example:

  Cpus_allowed:   ffffffff,ffffffff,ffffffff,ffffffff
  Mems_allowed:   ffffffff,ffffffff

Each cpuset is represented by a directory in the cpuset file system
containing the following files describing that cpuset:

 - cpus: list of CPUs in that cpuset
 - mems: list of Memory Nodes in that cpuset
 - memory_migrate flag: if set, move pages to cpusets nodes
 - cpu_exclusive flag: is cpu placement exclusive?
 - mem_exclusive flag: is memory placement exclusive?
 - tasks: list of tasks (by pid) attached to that cpuset
 - notify_on_release flag: run /sbin/cpuset_release_agent on exit?
 - memory_pressure: measure of how much paging pressure in cpuset

In addition, the root cpuset only has the following file:
 - memory_pressure_enabled flag: compute memory_pressure?

New cpusets are created using the mkdir system call or shell
command.  The properties of a cpuset, such as its flags, allowed
CPUs and Memory Nodes, and attached tasks, are modified by writing
to the appropriate file in that cpusets directory, as listed above.

The named hierarchical structure of nested cpusets allows partitioning
a large system into nested, dynamically changeable, "soft-partitions".

The attachment of each task, automatically inherited at fork by any
children of that task, to a cpuset allows organizing the work load
on a system into related sets of tasks such that each set is constrained
to using the CPUs and Memory Nodes of a particular cpuset.  A task
may be re-attached to any other cpuset, if allowed by the permissions
on the necessary cpuset file system directories.

Such management of a system "in the large" integrates smoothly with
the detailed placement done on individual tasks and memory regions
using the sched_setaffinity, mbind and set_mempolicy system calls.

The following rules apply to each cpuset:

 - Its CPUs and Memory Nodes must be a subset of its parents.
 - It can only be marked exclusive if its parent is.
 - If its cpu or memory is exclusive, they may not overlap any sibling.

These rules, and the natural hierarchy of cpusets, enable efficient
enforcement of the exclusive guarantee, without having to scan all
cpusets every time any of them change to ensure nothing overlaps a
exclusive cpuset.  Also, the use of a Linux virtual file system (vfs)
to represent the cpuset hierarchy provides for a familiar permission
and name space for cpusets, with a minimum of additional kernel code.

The cpus and mems files in the root (top_cpuset) cpuset are
read-only.  The cpus file automatically tracks the value of
cpu_online_map using a CPU hotplug notifier, and the mems file
automatically tracks the value of node_states[N_MEMORY]--i.e.,
nodes with memory--using the cpuset_track_online_nodes() hook.


1.4 What are exclusive cpusets ?
--------------------------------

If a cpuset is cpu or mem exclusive, no other cpuset, other than
a direct ancestor or descendent, may share any of the same CPUs or
Memory Nodes.

A cpuset that is cpu_exclusive has a scheduler (sched) domain
associated with it.  The sched domain consists of all CPUs in the
current cpuset that are not part of any exclusive child cpusets.
This ensures that the scheduler load balancing code only balances
against the CPUs that are in the sched domain as defined above and
not all of the CPUs in the system. This removes any overhead due to
load balancing code trying to pull tasks outside of the cpu_exclusive
cpuset only to be prevented by the tasks' cpus_allowed mask.

A cpuset that is mem_exclusive restricts kernel allocations for
page, buffer and other data commonly shared by the kernel across
multiple users.  All cpusets, whether mem_exclusive or not, restrict
allocations of memory for user space.  This enables configuring a
system so that several independent jobs can share common kernel data,
such as file system pages, while isolating each jobs user allocation in
its own cpuset.  To do this, construct a large mem_exclusive cpuset to
hold all the jobs, and construct child, non-mem_exclusive cpusets for
each individual job.  Only a small amount of typical kernel memory,
such as requests from interrupt handlers, is allowed to be taken
outside even a mem_exclusive cpuset.


1.5 What does notify_on_release do ?
------------------------------------

If the notify_on_release flag is enabled (1) in a cpuset, then whenever
the last task in the cpuset leaves (exits or attaches to some other
cpuset) and the last child cpuset of that cpuset is removed, then
the kernel runs the command /sbin/cpuset_release_agent, supplying the
pathname (relative to the mount point of the cpuset file system) of the
abandoned cpuset.  This enables automatic removal of abandoned cpusets.
The default value of notify_on_release in the root cpuset at system
boot is disabled (0).  The default value of other cpusets at creation
is the current value of their parents notify_on_release setting.


1.6 What is memory_pressure ?
-----------------------------
The memory_pressure of a cpuset provides a simple per-cpuset metric
of the rate that the tasks in a cpuset are attempting to free up in
use memory on the nodes of the cpuset to satisfy additional memory
requests.

This enables batch managers monitoring jobs running in dedicated
cpusets to efficiently detect what level of memory pressure that job
is causing.

This is useful both on tightly managed systems running a wide mix of
submitted jobs, which may choose to terminate or re-prioritize jobs that
are trying to use more memory than allowed on the nodes assigned them,
and with tightly coupled, long running, massively parallel scientific
computing jobs that will dramatically fail to meet required performance
goals if they start to use more memory than allowed to them.

This mechanism provides a very economical way for the batch manager
to monitor a cpuset for signs of memory pressure.  It's up to the
batch manager or other user code to decide what to do about it and
take action.

==> Unless this feature is enabled by writing "1" to the special file
    /dev/cpuset/memory_pressure_enabled, the hook in the rebalance
    code of __alloc_pages() for this metric reduces to simply noticing
    that the cpuset_memory_pressure_enabled flag is zero.  So only
    systems that enable this feature will compute the metric.

Why a per-cpuset, running average:

    Because this meter is per-cpuset, rather than per-task or mm,
    the system load imposed by a batch scheduler monitoring this
    metric is sharply reduced on large systems, because a scan of
    the tasklist can be avoided on each set of queries.

    Because this meter is a running average, instead of an accumulating
    counter, a batch scheduler can detect memory pressure with a
    single read, instead of having to read and accumulate results
    for a period of time.

    Because this meter is per-cpuset rather than per-task or mm,
    the batch scheduler can obtain the key information, memory
    pressure in a cpuset, with a single read, rather than having to
    query and accumulate results over all the (dynamically changing)
    set of tasks in the cpuset.

A per-cpuset simple digital filter (requires a spinlock and 3 words
of data per-cpuset) is kept, and updated by any task attached to that
cpuset, if it enters the synchronous (direct) page reclaim code.

A per-cpuset file provides an integer number representing the recent
(half-life of 10 seconds) rate of direct page reclaims caused by
the tasks in the cpuset, in units of reclaims attempted per second,
times 1000.


1.7 What is memory spread ?
---------------------------
There are two boolean flag files per cpuset that control where the
kernel allocates pages for the file system buffers and related in
kernel data structures.  They are called 'memory_spread_page' and
'memory_spread_slab'.

If the per-cpuset boolean flag file 'memory_spread_page' is set, then
the kernel will spread the file system buffers (page cache) evenly
over all the nodes that the faulting task is allowed to use, instead
of preferring to put those pages on the node where the task is running.

If the per-cpuset boolean flag file 'memory_spread_slab' is set,
then the kernel will spread some file system related slab caches,
such as for inodes and dentries evenly over all the nodes that the
faulting task is allowed to use, instead of preferring to put those
pages on the node where the task is running.

The setting of these flags does not affect anonymous data segment or
stack segment pages of a task.

By default, both kinds of memory spreading are off, and memory
pages are allocated on the node local to where the task is running,
except perhaps as modified by the tasks NUMA mempolicy or cpuset
configuration, so long as sufficient free memory pages are available.

When new cpusets are created, they inherit the memory spread settings
of their parent.

Setting memory spreading causes allocations for the affected page
or slab caches to ignore the tasks NUMA mempolicy and be spread
instead.    Tasks using mbind() or set_mempolicy() calls to set NUMA
mempolicies will not notice any change in these calls as a result of
their containing tasks memory spread settings.  If memory spreading
is turned off, then the currently specified NUMA mempolicy once again
applies to memory page allocations.

Both 'memory_spread_page' and 'memory_spread_slab' are boolean flag
files.  By default they contain "0", meaning that the feature is off
for that cpuset.  If a "1" is written to that file, then that turns
the named feature on.

The implementation is simple.

Setting the flag 'memory_spread_page' turns on a per-process flag
PF_SPREAD_PAGE for each task that is in that cpuset or subsequently
joins that cpuset.  The page allocation calls for the page cache
is modified to perform an inline check for this PF_SPREAD_PAGE task
flag, and if set, a call to a new routine cpuset_mem_spread_node()
returns the node to prefer for the allocation.

Similarly, setting 'memory_spread_cache' turns on the flag
PF_SPREAD_SLAB, and appropriately marked slab caches will allocate
pages from the node returned by cpuset_mem_spread_node().

The cpuset_mem_spread_node() routine is also simple.  It uses the
value of a per-task rotor cpuset_mem_spread_rotor to select the next
node in the current tasks mems_allowed to prefer for the allocation.

This memory placement policy is also known (in other contexts) as
round-robin or interleave.

This policy can provide substantial improvements for jobs that need
to place thread local data on the corresponding node, but that need
to access large file system data sets that need to be spread across
the several nodes in the jobs cpuset in order to fit.  Without this
policy, especially for jobs that might have one thread reading in the
data set, the memory allocation across the nodes in the jobs cpuset
can become very uneven.


1.8 How do I use cpusets ?
--------------------------

In order to minimize the impact of cpusets on critical kernel
code, such as the scheduler, and due to the fact that the kernel
does not support one task updating the memory placement of another
task directly, the impact on a task of changing its cpuset CPU
or Memory Node placement, or of changing to which cpuset a task
is attached, is subtle.

If a cpuset has its Memory Nodes modified, then for each task attached
to that cpuset, the next time that the kernel attempts to allocate
a page of memory for that task, the kernel will notice the change
in the tasks cpuset, and update its per-task memory placement to
remain within the new cpusets memory placement.  If the task was using
mempolicy MPOL_BIND, and the nodes to which it was bound overlap with
its new cpuset, then the task will continue to use whatever subset
of MPOL_BIND nodes are still allowed in the new cpuset.  If the task
was using MPOL_BIND and now none of its MPOL_BIND nodes are allowed
in the new cpuset, then the task will be essentially treated as if it
was MPOL_BIND bound to the new cpuset (even though its numa placement,
as queried by get_mempolicy(), doesn't change).  If a task is moved
from one cpuset to another, then the kernel will adjust the tasks
memory placement, as above, the next time that the kernel attempts
to allocate a page of memory for that task.

If a cpuset has its CPUs modified, then each task using that
cpuset does _not_ change its behavior automatically.  In order to
minimize the impact on the critical scheduling code in the kernel,
tasks will continue to use their prior CPU placement until they
are rebound to their cpuset, by rewriting their pid to the 'tasks'
file of their cpuset.  If a task had been bound to some subset of its
cpuset using the sched_setaffinity() call, and if any of that subset
is still allowed in its new cpuset settings, then the task will be
restricted to the intersection of the CPUs it was allowed on before,
and its new cpuset CPU placement.  If, on the other hand, there is
no overlap between a tasks prior placement and its new cpuset CPU
placement, then the task will be allowed to run on any CPU allowed
in its new cpuset.  If a task is moved from one cpuset to another,
its CPU placement is updated in the same way as if the tasks pid is
rewritten to the 'tasks' file of its current cpuset.

In summary, the memory placement of a task whose cpuset is changed is
updated by the kernel, on the next allocation of a page for that task,
but the processor placement is not updated, until that tasks pid is
rewritten to the 'tasks' file of its cpuset.  This is done to avoid
impacting the scheduler code in the kernel with a check for changes
in a tasks processor placement.

Normally, once a page is allocated (given a physical page
of main memory) then that page stays on whatever node it
was allocated, so long as it remains allocated, even if the
cpusets memory placement policy 'mems' subsequently changes.
If the cpuset flag file 'memory_migrate' is set true, then when
tasks are attached to that cpuset, any pages that task had
allocated to it on nodes in its previous cpuset are migrated
to the tasks new cpuset. The relative placement of the page within
the cpuset is preserved during these migration operations if possible.
For example if the page was on the second valid node of the prior cpuset
then the page will be placed on the second valid node of the new cpuset.

Also if 'memory_migrate' is set true, then if that cpusets
'mems' file is modified, pages allocated to tasks in that
cpuset, that were on nodes in the previous setting of 'mems',
will be moved to nodes in the new setting of 'mems.'
Pages that were not in the tasks prior cpuset, or in the cpusets
prior 'mems' setting, will not be moved.

There is an exception to the above.  If hotplug functionality is used
to remove all the CPUs that are currently assigned to a cpuset,
then the kernel will automatically update the cpus_allowed of all
tasks attached to CPUs in that cpuset to allow all CPUs.  When memory
hotplug functionality for removing Memory Nodes is available, a
similar exception is expected to apply there as well.  In general,
the kernel prefers to violate cpuset placement, over starving a task
that has had all its allowed CPUs or Memory Nodes taken offline.  User
code should reconfigure cpusets to only refer to online CPUs and Memory
Nodes when using hotplug to add or remove such resources.

There is a second exception to the above.  GFP_ATOMIC requests are
kernel internal allocations that must be satisfied, immediately.
The kernel may drop some request, in rare cases even panic, if a
GFP_ATOMIC alloc fails.  If the request cannot be satisfied within
the current tasks cpuset, then we relax the cpuset, and look for
memory anywhere we can find it.  It's better to violate the cpuset
than stress the kernel.

To start a new job that is to be contained within a cpuset, the steps are:

 1) mkdir /dev/cpuset
 2) mount -t cpuset none /dev/cpuset
 3) Create the new cpuset by doing mkdir's and write's (or echo's) in
    the /dev/cpuset virtual file system.
 4) Start a task that will be the "founding father" of the new job.
 5) Attach that task to the new cpuset by writing its pid to the
    /dev/cpuset tasks file for that cpuset.
 6) fork, exec or clone the job tasks from this founding father task.

For example, the following sequence of commands will setup a cpuset
named "Charlie", containing just CPUs 2 and 3, and Memory Node 1,
and then start a subshell 'sh' in that cpuset:

  mount -t cpuset none /dev/cpuset
  cd /dev/cpuset
  mkdir Charlie
  cd Charlie
  /bin/echo 2-3 > cpus
  /bin/echo 1 > mems
  /bin/echo $$ > tasks
  sh
  # The subshell 'sh' is now running in cpuset Charlie
  # The next line should display '/Charlie'
  cat /proc/self/cpuset

In the future, a C library interface to cpusets will likely be
available.  For now, the only way to query or modify cpusets is
via the cpuset file system, using the various cd, mkdir, echo, cat,
rmdir commands from the shell, or their equivalent from C.

The sched_setaffinity calls can also be done at the shell prompt using
SGI's runon or Robert Love's taskset.  The mbind and set_mempolicy
calls can be done at the shell prompt using the numactl command
(part of Andi Kleen's numa package).

2. Usage Examples and Syntax
============================

2.1 Basic Usage
---------------

Creating, modifying, using the cpusets can be done through the cpuset
virtual filesystem.

To mount it, type:
# mount -t cpuset none /dev/cpuset

Then under /dev/cpuset you can find a tree that corresponds to the
tree of the cpusets in the system. For instance, /dev/cpuset
is the cpuset that holds the whole system.

If you want to create a new cpuset under /dev/cpuset:
# cd /dev/cpuset
# mkdir my_cpuset

Now you want to do something with this cpuset.
# cd my_cpuset

In this directory you can find several files:
# ls
cpus  cpu_exclusive  mems  mem_exclusive  tasks

Reading them will give you information about the state of this cpuset:
the CPUs and Memory Nodes it can use, the processes that are using
it, its properties.  By writing to these files you can manipulate
the cpuset.

Set some flags:
# /bin/echo 1 > cpu_exclusive

Add some cpus:
# /bin/echo 0-7 > cpus

Add some mems:
# /bin/echo 0-7 > mems

Now attach your shell to this cpuset:
# /bin/echo $$ > tasks

You can also create cpusets inside your cpuset by using mkdir in this
directory.
# mkdir my_sub_cs

To remove a cpuset, just use rmdir:
# rmdir my_sub_cs
This will fail if the cpuset is in use (has cpusets inside, or has
processes attached).

2.2 Adding/removing cpus
------------------------

This is the syntax to use when writing in the cpus or mems files
in cpuset directories:

# /bin/echo 1-4 > cpus		-> set cpus list to cpus 1,2,3,4
# /bin/echo 1,2,3,4 > cpus	-> set cpus list to cpus 1,2,3,4

2.3 Setting flags
-----------------

The syntax is very simple:

# /bin/echo 1 > cpu_exclusive 	-> set flag 'cpu_exclusive'
# /bin/echo 0 > cpu_exclusive 	-> unset flag 'cpu_exclusive'

2.4 Attaching processes
-----------------------

# /bin/echo PID > tasks

Note that it is PID, not PIDs. You can only attach ONE task at a time.
If you have several tasks to attach, you have to do it one after another:

# /bin/echo PID1 > tasks
# /bin/echo PID2 > tasks
	...
# /bin/echo PIDn > tasks


3. Questions
============

Q: what's up with this '/bin/echo' ?
A: bash's builtin 'echo' command does not check calls to write() against
   errors. If you use it in the cpuset file system, you won't be
   able to tell whether a command succeeded or failed.

Q: When I attach processes, only the first of the line gets really attached !
A: We can only return one error code per call to write(). So you should also
   put only ONE pid.

4. Contact
==========

Web: http://www.bullopensource.org/cpuset