freebsd-skq/share/man/man4/numa.4
Mark Johnston b61f314290 Make it possible to disable NUMA support with a tunable.
This provides a chicken switch for anyone negatively impacted by
enabling NUMA in the amd64 GENERIC kernel configuration.  With
NUMA disabled at boot-time, information about the NUMA topology
is not exposed to the rest of the kernel, and all of physical
memory is viewed as coming from a single domain.

This method still has some performance overhead relative to disabling
NUMA support at compile time.

PR:		231460
Reviewed by:	alc, gallatin, kib
MFC after:	1 week
Sponsored by:	The FreeBSD Foundation
Differential Revision:	https://reviews.freebsd.org/D17439
2018-10-22 20:13:51 +00:00

151 lines
4.6 KiB
Groff

.\" Copyright (c) 2015 Adrian Chadd <adrian@FreeBSD.org>
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHORS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHORS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd October 22, 2018
.Dt NUMA 4
.Os
.Sh NAME
.Nm NUMA
.Nd Non-Uniform Memory Access
.Sh SYNOPSIS
.Cd options MAXMEMDOM
.Cd options NUMA
.Pp
.Sh DESCRIPTION
Non-Uniform Memory Access is a computer architecture design which
involves unequal costs between processors, memory and IO devices
in a given system.
.Pp
In a
.Nm
architecture, the latency to access specific memory or IO devices
depends upon which processor the memory or device is attached to.
Accessing memory local to a processor is faster than accessing memory
that is connected to one of the other processors.
.Fx
implements NUMA-aware memory allocation policies.
By default it attempts to ensure that allocations are balanced across
each domain.
Users may override the default domain selection policy using
.Xr cpuset 1 .
.Pp
.Nm
support is enabled when the
.Cd NUMA
option is specified in the kernel configuration file.
Each platform defines the
.Cd MAXMEMDOM
constant, which specifies the maximum number of supported NUMA domains.
This constant may be specified in the kernel configuration file.
.Nm
support can be disabled at boot time by setting the
.Va vm.numa.disabled
tunable to 1.
Other values for this tunable are currently ignored.
.Pp
Thread and process
.Nm
policies are controlled with the
.Xr cpuset_getdomain 2
and
.Xr cpuset_setdomain 2
syscalls.
The
.Xr cpuset 1
tool is available for starting processes with a non-default
policy, or to change the policy of an existing thread or process.
.Pp
Systems with non-uniform access to I/O devices may mark those devices
with the local VM domain identifier.
Drivers can find out their local domain information by calling
.Xr bus_get_domain 9 .
.Ss MIB Variables
The operation of
.Nm
is controlled and exposes information with these
.Xr sysctl 8
MIB variables:
.Pp
.Bl -tag -width indent -compact
.It Va vm.ndomains
The number of VM domains which have been detected.
.Pp
.It Va vm.phys_locality
A table indicating the relative cost of each VM domain to each other.
A value of 10 indicates equal cost.
A value of -1 means the locality map is not available or no
locality information is available.
.Pp
.It Va vm.phys_segs
The map of physical memory, grouped by VM domain.
.El
.Sh IMPLEMENTATION NOTES
The current
.Nm
implementation is VM-focused.
The hardware
.Nm
domains are mapped into a contiguous, non-sparse
VM domain space, starting from 0.
Thus, VM domain information (for example, the domain identifier) is not
necessarily the same as is found in the hardware specific information.
Policy information is available in both struct thread and struct proc.
.Sh SEE ALSO
.Xr cpuset 1 ,
.Xr cpuset_getaffinity 2 ,
.Xr cpuset_setaffinity 2 ,
.Xr bus_get_domain 9
.Sh HISTORY
.Nm
first appeared in
.Fx 9.0
as a first-touch allocation policy with a fail-over to round-robin allocation
and was not configurable.
It was then modified in
.Fx 10.0
to implement a round-robin allocation policy and was also not configurable.
.Pp
The
.Xr numa_getaffinity 2
and
.Xr numa_setaffinity 2
syscalls and the
.Xr numactl 1
tool first appeared in
.Fx 11.0
and were removed in
.Fx 12.0 .
The current implementation appeared in
.Fx 12.0 .
.Pp
.Sh AUTHORS
This manual page written by
.An Adrian Chadd Aq Mt adrian@FreeBSD.org .
.Sh NOTES
No statistics are kept to indicate how often
.Nm
allocation policies succeed or fail.