d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
337b148e02
freebsd-nq/lib/libc/sys/numa_getaffinity.2

198 lines
5.3 KiB
Groff
Raw Normal View History

Add an initial NUMA affinity/policy configuration for threads and processes. This is based on work done by jeff@ and jhb@, as well as the numa.diff patch that has been circulating when someone asks for first-touch NUMA on -10 or -11. * Introduce a simple set of VM policy and iterator types. * tie the policy types into the vm_phys path for now, mirroring how the initial first-touch allocation work was enabled. * add syscalls to control changing thread and process defaults. * add a global NUMA VM domain policy. * implement a simple cascade policy order - if a thread policy exists, use it; if a process policy exists, use it; use the default policy. * processes inherit policies from their parent processes, threads inherit policies from their parent threads. * add a simple tool (numactl) to query and modify default thread/process policities. * add documentation for the new syscalls, for numa and for numactl. * re-enable first touch NUMA again by default, as now policies can be set in a variety of methods. This is only relevant for very specific workloads. This doesn't pretend to be a final NUMA solution. The previous defaults in -HEAD (with MAXMEMDOM set) can be achieved by 'sysctl vm.default_policy=rr'. This is only relevant if MAXMEMDOM is set to something other than 1. Ie, if you're using GENERIC or a modified kernel with non-NUMA, then this is a glorified no-op for you. Thank you to Norse Corp for giving me access to rather large (for FreeBSD!) NUMA machines in order to develop and verify this. Thank you to Dell for providing me with dual socket sandybridge and westmere v3 hardware to do NUMA development with. Thank you to Scott Long at Netflix for providing me with access to the two-socket, four-domain haswell v3 hardware. Thank you to Peter Holm for running the stress testing suite against the NUMA branch during various stages of development! Tested: * MIPS (regression testing; non-NUMA) * i386 (regression testing; non-NUMA GENERIC) * amd64 (regression testing; non-NUMA GENERIC) * westmere, 2 socket (thankyou norse!) * sandy bridge, 2 socket (thankyou dell!) * ivy bridge, 2 socket (thankyou norse!) * westmere-EX, 4 socket / 1TB RAM (thankyou norse!) * haswell, 2 socket (thankyou norse!) * haswell v3, 2 socket (thankyou dell) * haswell v3, 2x18 core (thankyou scott long / netflix!) * Peter Holm ran a stress test suite on this work and found one issue, but has not been able to verify it (it doesn't look NUMA related, and he only saw it once over many testing runs.) * I've tested bhyve instances running in fixed NUMA domains and cpusets; all seems to work correctly. Verified: * intel-pcm - pcm-numa.x and pcm-memory.x, whilst selecting different NUMA policies for processes under test. Review: This was reviewed through phabricator (https://reviews.freebsd.org/D2559) as well as privately and via emails to freebsd-arch@. The git history with specific attributes is available at https://github.com/erikarn/freebsd/ in the NUMA branch (https://github.com/erikarn/freebsd/compare/local/adrian_numa_policy). This has been reviewed by a number of people (stas, rpaulo, kib, ngie, wblock) but not achieved a clear consensus. My hope is that with further exposure and testing more functionality can be implemented and evaluated. Notes: * The VM doesn't handle unbalanced domains very well, and if you have an overly unbalanced memory setup whilst under high memory pressure, VM page allocation may fail leading to a kernel panic. This was a problem in the past, but it's much more easily triggered now with these tools. * This work only controls the path through vm_phys; it doesn't yet strongly/predictably affect contigmalloc, KVA placement, UMA, etc. So, driver placement of memory isn't really guaranteed in any way. That's next on my plate. Sponsored by: Norse Corp, Inc.; Dell
2015-07-11 15:21:37 +00:00
.\" Copyright (c) 2008 Christian Brueffer
.\" Copyright (c) 2008 Jeffrey Roberson
.\" Copyright (c) 2015 Adrian Chadd
.\" 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 AUTHOR 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 AUTHOR 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 May 7, 2015
.Dt NUMA_GETAFFINITY 2
.Os
.Sh NAME
.Nm numa_getaffinity ,
.Nm numa_setaffinity
.Nd manage NUMA affinity
.Sh LIBRARY
.Lb libc
.Sh SYNOPSIS
.In sys/param.h
.In sys/numa.h
.Ft int
.Fn numa_getaffinity "cpuwhich_t which" "id_t id" "struct vm_domain_policy_entry *policy"
.Ft int
.Fn numa_setaffinity "cpuwhich_t which" "id_t id" "const struct vm_domain_policy_entry *policy"
.Sh DESCRIPTION
.Fn numa_getaffinity
and
.Fn numa_setaffinity
allow the manipulation of NUMA policies available to processes and threads.
These functions may manipulate NUMA policies that contain many processes
or affect only a single object.
.Pp
Valid values for the
.Fa which
argument are documented in
.Xr cpuset 2 .
These arguments specify which object set are used.
Only
.Dv CPU_WHICH_TID
and
.Dv CPU_WHICH_PID
can be manipulated.
.Pp
The
.Fa policy
entry contains a vm_domain_policy_entry with the following fields:
.Bd -literal
struct vm_domain_policy_entry {
vm_domain_policy_type_t policy; /* VM policy */
int domain; /* VM domain, if applicable */
}
.Ed
.Fa vm_domain_policy_type_t policy
is one these:
.Bl -tag -width VM_POLICY_NONE
.It Dv VM_POLICY_NONE
Reset the domain back to none.
Any parent object NUMA domain policy will apply.
The only valid value for
.Dv domain
is -1.
.It Dv VM_POLICY_ROUND_ROBIN
Select round-robin policy.
Pages will be allocated round-robin from each VM domain in order.
The only valid value for
.Dv domain
is -1.
.It Dv VM_POLICY_FIXED_DOMAIN
Select fixed-domain only policy.
Pages will be allocated from the given
.Dv domain
which must be set to a valid VM domain.
Pages will not be allocated from another domain if
.Dv domain
is out of free pages.
.It Dv VM_POLICY_FIXED_DOMAIN_ROUND_ROBIN
Select fixed-domain only policy.
Pages will be allocated from
.Dv domain
which must be set to a valid VM domain.
If page allocation fails, pages will be round-robin
allocated from another domain if
.Dv domain
is out of free pages.
.It Dv VM_POLICY_FIRST_TOUCH
Select first-touch policy.
Pages will be allocated from the NUMA domain which the thread
is currently scheduled upon.
Pages will not be allocated from another domain if the current domain
is out of free pages.
The only valid value for
.Dv domain
is -1.
.It Dv VM_POLICY_FIRST_TOUCH_ROUND_ROBIN
Select first-touch policy.
Pages will be allocated from the NUMA domain which the thread
is currently scheduled upon.
Pages will be allocated round-robin from another domain if the
current domain is out of free pages.
The only valid value for
.Dv domain
is -1.
.El
.Pp
Note that the VM might assign some pages from other domains.
For example, if an existing page allocation is covered by a superpage
allocation.
.Pp
.Fn numa_getaffinity
retrieves the
NUMA policy from the object specified by
.Fa which
and
.Fa id
and stores it in the space provided by
.Fa policy .
.Pp
.Fn numa_setaffinity
attempts to set the NUMA policy for the object specified by
.Fa which
and
.Fa id
to the policy in
.Fa policy .
.Sh RETURN VALUES
.Rv -std
.Sh ERRORS
.Va errno
can contain these error codes:
.Bl -tag -width Er
.It Bq Er EINVAL
The
.Fa level
or
.Fa which
argument was not a valid value.
.It Bq Er EINVAL
The
.Fa policy
argument specified when calling
.Fn numa_setaffinity
did not contain a valid policy.
.It Bq Er EFAULT
The policy pointer passed was invalid.
.It Bq Er ESRCH
The object specified by the
.Fa id
and
.Fa which
arguments could not be found.
.It Bq Er ERANGE
The
.Fa domain
in the given policy
was out of the range of possible VM domains available.
.It Bq Er EPERM
The calling process did not have the credentials required to complete the
operation.
.El
.Sh SEE ALSO
.Xr cpuset 1 ,
.Xr numactl 1 ,
.Xr cpuset 2 ,
.Xr cpuset_getaffinity 2 ,
.Xr cpuset_getid 2 ,
.Xr cpuset_setaffinity 2 ,
.Xr cpuset_setid 2 ,
.Xr pthread_affinity_np 3 ,
.Xr pthread_attr_affinity_np 3 ,
.Xr numa 4
.Sh HISTORY
The
.Nm
family of system calls first appeared in
.Fx 11.0 .
.Sh AUTHORS
.An Adrian Chadd Aq Mt adrian@FreeBSD.org
Reference in New Issue Copy Permalink