Commit bd5e09cf authored by Paul Jackson's avatar Paul Jackson Committed by Linus Torvalds

[PATCH] cpuset: document additional features

Document the additional cpuset features:
	notify_on_release
	marker_pid
	memory_pressure
	memory_pressure_enabled

Rearrange and improve formatting of existing documentation for
cpu_exclusive and mem_exclusive features.
Signed-off-by: default avatarPaul Jackson <pj@sgi.com>
Signed-off-by: default avatarAndrew Morton <akpm@osdl.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@osdl.org>
parent 3e0d98b9
...@@ -14,7 +14,11 @@ CONTENTS: ...@@ -14,7 +14,11 @@ CONTENTS:
1.1 What are cpusets ? 1.1 What are cpusets ?
1.2 Why are cpusets needed ? 1.2 Why are cpusets needed ?
1.3 How are cpusets implemented ? 1.3 How are cpusets implemented ?
1.4 How do I use cpusets ? 1.4 What are exclusive cpusets ?
1.5 What does notify_on_release do ?
1.6 What is a marker_pid ?
1.7 What is memory_pressure ?
1.8 How do I use cpusets ?
2. Usage Examples and Syntax 2. Usage Examples and Syntax
2.1 Basic Usage 2.1 Basic Usage
2.2 Adding/removing cpus 2.2 Adding/removing cpus
...@@ -49,29 +53,6 @@ its cpus_allowed vector, and the kernel page allocator will not ...@@ -49,29 +53,6 @@ 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 allocate a page on a node that is not allowed in the requesting tasks
mems_allowed vector. mems_allowed vector.
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 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 balacing 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.
User level code may create and destroy cpusets by name in the cpuset User level code may create and destroy cpusets by name in the cpuset
virtual file system, manage the attributes and permissions of these virtual file system, manage the attributes and permissions of these
cpusets and which CPUs and Memory Nodes are assigned to each cpuset, cpusets and which CPUs and Memory Nodes are assigned to each cpuset,
...@@ -196,6 +177,12 @@ containing the following files describing that cpuset: ...@@ -196,6 +177,12 @@ containing the following files describing that cpuset:
- cpu_exclusive flag: is cpu placement exclusive? - cpu_exclusive flag: is cpu placement exclusive?
- mem_exclusive flag: is memory placement exclusive? - mem_exclusive flag: is memory placement exclusive?
- tasks: list of tasks (by pid) attached to that cpuset - tasks: list of tasks (by pid) attached to that cpuset
- notify_on_release flag: run /sbin/cpuset_release_agent on exit?
- marker_pid: pid of user task in co-ordinated operation sequence
- 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 New cpusets are created using the mkdir system call or shell
command. The properties of a cpuset, such as its flags, allowed command. The properties of a cpuset, such as its flags, allowed
...@@ -229,7 +216,148 @@ exclusive cpuset. Also, the use of a Linux virtual file system (vfs) ...@@ -229,7 +216,148 @@ exclusive cpuset. Also, the use of a Linux virtual file system (vfs)
to represent the cpuset hierarchy provides for a familiar permission to represent the cpuset hierarchy provides for a familiar permission
and name space for cpusets, with a minimum of additional kernel code. and name space for cpusets, with a minimum of additional kernel code.
1.4 How do I use cpusets ?
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 a marker_pid ?
--------------------------
The marker_pid helps manage cpuset changes safely from user space.
The interface presented to user space for cpusets uses system wide
numbering of CPUs and Memory Nodes. It is the responsibility of
user level code, presumably in a library, to present cpuset-relative
numbering to applications when that would be more useful to them.
However if a task is moved to a different cpuset, or if the 'cpus' or
'mems' of a cpuset are changed, then we need a way for such library
code to detect that its cpuset-relative numbering has changed, when
expressed using system wide numbering.
The kernel cannot safely allow user code to lock kernel resources.
The kernel could deliver out-of-band notice of cpuset changes by
such mechanisms as signals or usermodehelper callbacks, however
this can't be synchronously delivered to library code linked in
applications without intruding on the IPC mechanisms available to
the app. The kernel could require user level code to do all the work,
tracking the cpuset state before and during changes, to verify no
unexpected change occurred, but this becomes an onerous task.
The "marker_pid" cpuset field provides a simple way to make this task
less onerous on user library code. A task writes its pid to a cpusets
"marker_pid" at the start of a sequence of queries and updates,
and check as it goes that the cpusets marker_pid doesn't change.
The pread(2) system call does a seek and read in a single call.
If the marker_pid changes, the user code should retry the required
sequence of operations.
Anytime that a task modifies the "cpus" or "mems" of a cpuset,
unless it's pid is in the cpusets marker_pid field, the kernel zeros
this field.
The above was inspired by the load linked and store conditional
(ll/sc) instructions in the MIPS II instruction set.
1.7 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.
1.8 How do I use cpusets ?
-------------------------- --------------------------
In order to minimize the impact of cpusets on critical kernel In order to minimize the impact of cpusets on critical kernel
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment