It can useful for code outside the VM system to look up the NUMA domain
of a page backing a virtual or physical address, specifically when
creating NUMA-aware data structures. We have _vm_phys_domain() for
this, but the leading underscore implies that it's an internal function,
and vm_phys.h has dependencies on a number of other headers.
Rename vm_phys_domain() to vm_page_domain(), and _vm_phys_domain() to
vm_phys_domain(). Make the latter an inline function.
Add _vm_phys.h and define struct vm_phys_seg there so that it's easier
to use in other headers. Include it from vm_page.h so that
vm_page_domain() can be defined there.
Include machine/vmparam.h from _vm_phys.h since it depends directly on
some constants defined there.
Reviewed by: alc
Reviewed by: dougm, kib (earlier versions)
Differential Revision: https://reviews.freebsd.org/D27207
- Add vm_phys_early_add_seg(), complementing vm_phys_early_alloc(), to
ensure that segments registered during hammer_time() are placed in the
right domain. Otherwise, since the SRAT is not parsed at that point,
we just add them to domain 0, which may be incorrect and results in a
domain with only several MB worth of memory.
- Fix uma_startup1() to try allocating memory for zones from any domain.
If domain 0 is unpopulated, the allocation will simply fail, resulting
in a page fault slightly later during boot.
- Change _vm_phys_domain() to return -1 for addresses not covered by the
affinity table, and change vm_phys_early_alloc() to handle wildcard
domains. This is necessary on amd64, where the page array is dense
and pmap_page_array_startup() may allocate page table pages for
non-existent page frames.
Reported and tested by: Rafael Kitover <rkitover@gmail.com>
Reviewed by: cem (earlier version), kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D25001
r357614 added CTLFLAG_NEEDGIANT to make it easier to find nodes that are
still not MPSAFE (or already are but aren’t properly marked).
Use it in preparation for a general review of all nodes.
This is non-functional change that adds annotations to SYSCTL_NODE and
SYSCTL_PROC nodes using one of the soon-to-be-required flags.
Mark all obvious cases as MPSAFE. All entries that haven't been marked
as MPSAFE before are by default marked as NEEDGIANT
Approved by: kib (mentor, blanket)
Commented by: kib, gallatin, melifaro
Differential Revision: https://reviews.freebsd.org/D23718
The macro RB_INITIALIZER ignores its argument, but is documented to
require "&head" as argument to initialize "head". So using
"_vm_phys_fictitious_tree" as the argument to initialize
"vm_phys_fictitious_tree" is an inconsequential error, corrected here.
Discussed with: alc
NUMA aware boot time memory allocator that will be used to allocate early
domain correct structures. Code partially submitted by gallatin.
Reviewed by: gallatin, kib
Tested by: pho
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D21251
doing so adds more flexibility with less redundant code.
Reviewed by: jhb, markj, kib
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D21250
- UMA_XDOMAIN enables an additional per-cpu bucket for freed memory that
was freed on a different domain from where it was allocated. This is
only used for UMA_ZONE_NUMA (first-touch) zones.
- UMA_FIRSTTOUCH sets the default UMA policy to be first-touch for all
zones. This tries to maintain locality for kernel memory.
Reviewed by: gallatin, alc, kib
Tested by: pho, gallatin
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D20929
power-of-two page block it frees, launching an unsuccessful search for
a buddy to pair up with each time. The only possible buddy-up mergers
are across the boundaries of the freed region, so change
vm_phys_free_contig simply to enqueue the freed interior blocks, via a
new function vm_phys_enqueue_contig, and then call vm_phys_free_pages
on the bounding blocks to create as big a cross-boundary block as
possible after buddy-merging.
The only callers of vm_phys_free_contig at the moment call it in
situations where merging blocks across the boundary is clearly
impossible, so just call vm_phys_enqueue_contig in those places and
avoid trying to buddy-up at all.
One beneficiary of this change is in breaking reservations. For the
case where memory is freed in breaking a reservation with only the
first and last pages allocated, the number of cycles consumed by the
operation drops about 11% with this change.
Suggested by: alc
Reviewed by: alc
Approved by: kib, markj (mentors)
Differential Revision: https://reviews.freebsd.org/D16901
In order to allow single kernel to use PAE pagetables on i386 if
hardware supports it, and fall back to classic two-level paging
structures if not, superpage code should be able to adopt to either 2M
or 4M superpages size. There I make MI VM structures large enough to
track the biggest possible superpage, by allowing architecture to
define VM_NFREEORDER_MAX and VM_LEVEL_0_ORDER_MAX constants.
Corresponding VM_NFREEORDER and VM_LEVEL_0_ORDER symbols can be
defined as runtime values and must be less than the _MAX constants.
If architecture does not define _MAXs, it is assumed that _MAX ==
normal constant.
Reviewed by: markj
Tested by: pho (as part of the larger patch)
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D18853
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
Pre-defined policies are useful when integrating the domainset(9)
policy machinery into various kernel memory allocators.
The refactoring will make it easier to add NUMA support for other
architectures.
No functional change intended.
Reviewed by: alc, gallatin, jeff, kib
Tested by: pho (part of a larger patch)
MFC after: 3 days
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D17416
Use these predicates instead of inline references to vm_min_domains.
Also add a global all_domains set, akin to all_cpus.
Reviewed by: alc, jeff, kib
Approved by: re (gjb)
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D17278
that can be coalesced. To be clear, fragmentation of phys_avail[] is not
the cause. This fragmentation of vm_phys_segs[] arises from the "special"
calls to vm_phys_add_seg(), in other words, not those that derive directly
from phys_avail[], but those that we create for the initial kernel page
table pages and now for the kernel and modules loaded at boot time. Since
we sometimes iterate over the physical memory segments, coalescing these
segments at initialization time is a worthwhile change.
Reviewed by: kib, markj
Approved by: re (rgrimes)
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D16976
There's no differene between VM_FREELIST_ISADMA and VM_FREELIST_LOWMEM
except for the default boundary (16MB on x86 and 256MB on MIPS, but
they are otherwise the same). We don't need both for any system we
support (there were some really old ARC systems that did have ISA/EISA
bus, but we never ran on them and they are too old to ever grow
support for).
Differential Review: https://reviews.freebsd.org/D16290
and vm_phys_alloc_seg_contig() instead of vm_phys_free_contig(). In
short, vm_phys_enq_range() is simpler and faster than the more general
vm_phys_free_contig(), and in the case of vm_phys_alloc_seg_contig(),
vm_phys_free_contig() was placing the excess physical pages at the
wrong end of the queues.
In collaboration with: Doug Moore <dougm@rice.edu>
1. Optimize the order computation.
2. Update the pool for all of the chunks that are removed from the free
page lists, and not just the first chunk.
3. Simplify the code for returning excess pages to the free page lists.
Reviewed by: Doug Moore <dougm@rice.edu>
that it does not cause rapid fragmentation of the free physical memory.
Reviewed by: jeff, markj (an earlier version)
Differential Revision: https://reviews.freebsd.org/D15976
Currently both the page lock and a page queue lock must be held in
order to enqueue, dequeue or requeue a page in a given page queue.
The queue locks are a scalability bottleneck in many workloads. This
change reduces page queue lock contention by batching queue operations.
To detangle the page and page queue locks, per-CPU batch queues are
used to reference pages with pending queue operations. The requested
operation is encoded in the page's aflags field with the page lock
held, after which the page is enqueued for a deferred batch operation.
Page queue scans are similarly optimized to minimize the amount of
work performed with a page queue lock held.
Reviewed by: kib, jeff (previous versions)
Tested by: pho
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D14893
per-cpu alloc and free of pages. The cache is filled with as few trips
to the phys allocator as possible by the use of a new
vm_phys_alloc_npages() function which allocates as many as N pages.
This code was originally by markj with the import function rewritten by
me.
Reviewed by: markj, kib
Tested by: pho
Sponsored by: Netflix, Dell/EMC Isilon
Differential Revision: https://reviews.freebsd.org/D14905
It is possible to provide insane values for size in contigmalloc(9)
request, which usually not reaches the phys allocator due to failing
KVA allocation. But with the forthcoming 4/4 i386, where 32bit
architecture has almost 4G KVA, contigmalloc(1G) is not unreasonable
outright and KVA might be available sometimes.
Then, the calculation of pa_end could wrap around, depending on the
physical address, and the checks in vm_phys_alloc_seg_contig() would
pass while the iteration in the loop after the 'done' label goes out
of the vm_page_array bounds.
Fix it by detecting the wrap.
Reported and tested by: pho
Reviewed by: alc, markj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
Differential revision: https://reviews.freebsd.org/D14767
global to per-domain state. Protect reservations with the free lock
from the domain that they belong to. Refactor to make vm domains more
of a first class object.
Reviewed by: markj, kib, gallatin
Tested by: pho
Sponsored by: Netflix, Dell/EMC Isilon
Differential Revision: https://reviews.freebsd.org/D14000
allocated with a tag to come from the specified domain if it meets the
other constraints provided by the tag. Automatically create a tag at
the root of each bus specifying the domain local to that bus if
available.
Reviewed by: jhb, kib
Tested by: pho
Sponsored by: Netflix, Dell/EMC Isilon
Differential Revision: https://reviews.freebsd.org/D13545
userspace to control NUMA policy administratively and programmatically.
Implement domainset based iterators in the page layer.
Remove the now legacy numa_* syscalls.
Cleanup some header polution created by having seq.h in proc.h.
Reviewed by: markj, kib
Discussed with: alc
Tested by: pho
Sponsored by: Netflix, Dell/EMC Isilon
Differential Revision: https://reviews.freebsd.org/D13403
Commit r326346 moved domain iterators from physical layer to vm_page one,
but it also removed translation of freelist to flind for
vm_page_alloc_freelist() call. Before it expects VM_FREELIST_ parameter,
but after it expect freelist index.
On small WiFi boxes with few megabytes of RAM, there is only one freelist
VM_FREELIST_LOWMEM (1) and there is no VM_FREELIST_DEFAULT(0) (see file
sys/mips/include/vmparam.h). It results in freelist 1 with flind 0.
At first, this commit renames flind to freelist in vm_page_alloc_freelist
to avoid misunderstanding about input parameters. Then on physical layer it
restores translation for correct handling of freelist parameter.
Reported by: landonf
Reviewed by: jeff
Differential Revision: https://reviews.freebsd.org/D13351
This gives a marginal improvement in the vm_page_array initialization
time. Also garbage-collect the now-unused vm_phys_paddr_to_segind().
Reviewed by: alc, kib
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D13270
Mainly focus on files that use BSD 2-Clause license, however the tool I
was using misidentified many licenses so this was mostly a manual - error
prone - task.
The Software Package Data Exchange (SPDX) group provides a specification
to make it easier for automated tools to detect and summarize well known
opensource licenses. We are gradually adopting the specification, noting
that the tags are considered only advisory and do not, in any way,
superceed or replace the license texts.
No functional change intended.
We currently initialize the vm_page array in three passes: one to zero
the array, one to initialize the "order" field of each page (necessary
when inserting them into the vm_phys buddy allocator one-by-one), and
one to initialize the remaining non-zero fields and individually insert
each page into the allocator.
Merge the three passes into one following a suggestion from alc:
initialize vm_page fields in a single pass, and use vm_phys_free_contig()
to efficiently insert physical memory segments into the buddy allocator.
This reduces the initialization time to a third or a quarter of what it
was before on most systems that I tested.
Reviewed by: alc, kib
MFC after: 3 weeks
Differential Revision: https://reviews.freebsd.org/D12248
Idle page zeroing has been disabled by default on all architectures since
r170816 and has some bugs that make it seemingly unusable. Specifically,
the idle-priority pagezero thread exacerbates contention for the free page
lock, and yields the CPU without releasing it in non-preemptive kernels. The
pagezero thread also does not behave correctly when superpage reservations
are enabled: its target is a function of v_free_count, which includes
reserved-but-free pages, but it is only able to zero pages belonging to the
physical memory allocator.
Reviewed by: alc, imp, kib
Differential Revision: https://reviews.freebsd.org/D7714
rounddown2 tends to produce longer lines than the original code
and when the code has a high indentation level it was not really
advantageous to do the replacement.
This tries to strike a balance between readability using the macros
and flexibility of having the expressions, so not everything is
converted.
VM_NUMA_ALLOC is used to enable use of domain-aware memory allocation in
the virtual memory system. DEVICE_NUMA is used to enable affinity
reporting for devices such as bus_get_domain().
MAXMEMDOM must still be set to a value greater than for any NUMA support
to be effective. Note that 'cpuset -gd' always works if MAXMEMDOM is
enabled and the system supports NUMA.
Reviewed by: kib
Differential Revision: https://reviews.freebsd.org/D5782
than ascending order in vm_phys_alloc_contig() so that, for example, a
sequence of contigmalloc(low=0, high=4GB) calls doesn't exhaust the supply
of low physical memory resulting in a later contigmalloc(low=0, high=1MB)
failure.
Reported by: cy
Tested by: cy
Sponsored by: EMC / Isilon Storage Division
address and use this mechanism when:
1. kmem_alloc_{attr,contig}() can't find suitable free pages in the physical
memory allocator's free page lists. This replaces the long-standing
approach of scanning the inactive and inactive queues, converting clean
pages into PG_CACHED pages and laundering dirty pages. In contrast, the
new mechanism does not use PG_CACHED pages nor does it trigger a large
number of I/O operations.
2. on 32-bit MIPS processors, uma_small_alloc() and the pmap can't find
free pages in the physical memory allocator's free page lists that are
covered by the direct map. Tested by: adrian
3. ttm_bo_global_init() and ttm_vm_page_alloc_dma32() can't find suitable
free pages in the physical memory allocator's free page lists.
In the coming months, I expect that this new mechanism will be applied in
other places. For example, balloon drivers should use relocation to
minimize fragmentation of the guest physical address space.
Make vm_phys_alloc_contig() a little smarter (and more efficient in some
cases). Specifically, use vm_phys_segs[] earlier to avoid scanning free
page lists that can't possibly contain suitable pages.
Reviewed by: kib, markj
Glanced at: jhb
Discussed with: jeff
Sponsored by: EMC / Isilon Storage Division
Differential Revision: https://reviews.freebsd.org/D4444
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
a basic ACPI SLIT table parser.
For now this just exports the map via sysctl; it'll eventually be useful
to userland when there's more useful NUMA support in -HEAD.
* Add an optional mem_locality map;
* add a mapping function taking from/to domain and returning the
relative cost, or -1 if it's not available;
* Add a very basic SLIT parser to x86 ACPI.
Differential Revision: https://reviews.freebsd.org/D2460
Reviewed by: rpaulo, stas, jhb
Sponsored by: Norse Corp, Inc (hardware, coding); Dell (hardware)
