linux/Documentation/vm
Andrea Arcangeli 2c653d0ee2 ksm: introduce ksm_max_page_sharing per page deduplication limit
Without a max deduplication limit for each KSM page, the list of the
rmap_items associated to each stable_node can grow infinitely large.

During the rmap walk each entry can take up to ~10usec to process
because of IPIs for the TLB flushing (both for the primary MMU and the
secondary MMUs with the MMU notifier).  With only 16GB of address space
shared in the same KSM page, that would amount to dozens of seconds of
kernel runtime.

A ~256 max deduplication factor will reduce the latencies of the rmap
walks on KSM pages to order of a few msec.  Just doing the
cond_resched() during the rmap walks is not enough, the list size must
have a limit too, otherwise the caller could get blocked in (schedule
friendly) kernel computations for seconds, unexpectedly.

There's room for optimization to significantly reduce the IPI delivery
cost during the page_referenced(), but at least for page_migration in
the KSM case (used by hard NUMA bindings, compaction and NUMA balancing)
it may be inevitable to send lots of IPIs if each rmap_item->mm is
active on a different CPU and there are lots of CPUs.  Even if we ignore
the IPI delivery cost, we've still to walk the whole KSM rmap list, so
we can't allow millions or billions (ulimited) number of entries in the
KSM stable_node rmap_item lists.

The limit is enforced efficiently by adding a second dimension to the
stable rbtree.  So there are three types of stable_nodes: the regular
ones (identical as before, living in the first flat dimension of the
stable rbtree), the "chains" and the "dups".

Every "chain" and all "dups" linked into a "chain" enforce the invariant
that they represent the same write protected memory content, even if
each "dup" will be pointed by a different KSM page copy of that content.
This way the stable rbtree lookup computational complexity is unaffected
if compared to an unlimited max_sharing_limit.  It is still enforced
that there cannot be KSM page content duplicates in the stable rbtree
itself.

Adding the second dimension to the stable rbtree only after the
max_page_sharing limit hits, provides for a zero memory footprint
increase on 64bit archs.  The memory overhead of the per-KSM page
stable_tree and per virtual mapping rmap_item is unchanged.  Only after
the max_page_sharing limit hits, we need to allocate a stable_tree
"chain" and rb_replace() the "regular" stable_node with the newly
allocated stable_node "chain".  After that we simply add the "regular"
stable_node to the chain as a stable_node "dup" by linking hlist_dup in
the stable_node_chain->hlist.  This way the "regular" (flat) stable_node
is converted to a stable_node "dup" living in the second dimension of
the stable rbtree.

During stable rbtree lookups the stable_node "chain" is identified as
stable_node->rmap_hlist_len == STABLE_NODE_CHAIN (aka
is_stable_node_chain()).

When dropping stable_nodes, the stable_node "dup" is identified as
stable_node->head == STABLE_NODE_DUP_HEAD (aka is_stable_node_dup()).

The STABLE_NODE_DUP_HEAD must be an unique valid pointer never used
elsewhere in any stable_node->head/node to avoid a clashes with the
stable_node->node.rb_parent_color pointer, and different from
&migrate_nodes.  So the second field of &migrate_nodes is picked and
verified as always safe with a BUILD_BUG_ON in case the list_head
implementation changes in the future.

The STABLE_NODE_DUP is picked as a random negative value in
stable_node->rmap_hlist_len.  rmap_hlist_len cannot become negative when
it's a "regular" stable_node or a stable_node "dup".

The stable_node_chain->nid is irrelevant.  The stable_node_chain->kpfn
is aliased in a union with a time field used to rate limit the
stable_node_chain->hlist prunes.

The garbage collection of the stable_node_chain happens lazily during
stable rbtree lookups (as for all other kind of stable_nodes), or while
disabling KSM with "echo 2 >/sys/kernel/mm/ksm/run" while collecting the
entire stable rbtree.

While the "regular" stable_nodes and the stable_node "dups" must wait
for their underlying tree_page to be freed before they can be freed
themselves, the stable_node "chains" can be freed immediately if the
stable_node->hlist turns empty.  This is because the "chains" are never
pointed by any page->mapping and they're effectively stable rbtree KSM
self contained metadata.

[akpm@linux-foundation.org: fix non-NUMA build]
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Tested-by: Petr Holasek <pholasek@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Evgheni Dereveanchin <ederevea@redhat.com>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Gavin Guo <gavin.guo@canonical.com>
Cc: Jay Vosburgh <jay.vosburgh@canonical.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-07-06 16:24:31 -07:00
..
.gitignore
00-INDEX Documentation: vm, add hugetlbfs reservation overview 2017-05-03 15:52:11 -07:00
active_mm.txt
balance mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd 2015-11-06 17:50:42 -08:00
cleancache.txt cleancache: forbid overriding cleancache_ops 2015-04-14 16:49:03 -07:00
frontswap.txt
highmem.txt
hugetlbfs_reserv.txt Documentation: vm, add hugetlbfs reservation overview 2017-05-03 15:52:11 -07:00
hugetlbpage.txt Documentation: vm: Spelling s/paltform/platform/g 2016-05-14 10:15:10 -06:00
hwpoison.txt mm/memory-failure.c: support use of a dedicated thread to handle SIGBUS(BUS_MCEERR_AO) 2014-06-04 16:54:13 -07:00
idle_page_tracking.txt mm: introduce idle page tracking 2015-09-10 13:29:01 -07:00
ksm.txt ksm: introduce ksm_max_page_sharing per page deduplication limit 2017-07-06 16:24:31 -07:00
numa docs: fix locations of several documents that got moved 2016-10-24 08:12:35 -02:00
numa_memory_policy.txt Documenation: update cgroup's document path 2016-08-03 15:43:58 -06:00
overcommit-accounting mm: add overcommit_kbytes sysctl variable 2014-01-21 16:19:44 -08:00
page_frags mm: add documentation for page fragment APIs 2017-01-10 18:31:55 -08:00
page_migration Three fixes for the docs build, including removing an annoying warning on 2016-08-07 10:23:17 -04:00
page_owner.txt mm, page_owner: convert page_owner_inited to static key 2016-03-15 16:55:16 -07:00
pagemap.txt Documentation typo: wrong page flag bit for KPF_HUGE 2016-04-15 15:47:10 -06:00
remap_file_pages.txt mm: replace remap_file_pages() syscall with emulation 2015-02-10 14:30:30 -08:00
slub.txt slub: convert SLAB_DEBUG_FREE to SLAB_CONSISTENCY_CHECKS 2016-03-15 16:55:16 -07:00
soft-dirty.txt mm: track vma changes with VM_SOFTDIRTY bit 2013-09-11 15:57:56 -07:00
split_page_table_lock mm: make compound_head() robust 2015-11-06 17:50:42 -08:00
transhuge.txt Documentation/vm/transhuge.txt: fix trivial typos 2017-05-08 17:15:14 -07:00
unevictable-lru.txt Three fixes for the docs build, including removing an annoying warning on 2016-08-07 10:23:17 -04:00
userfaultfd.txt userfaultfd: non-cooperative: rollback userfaultfd_exit 2017-03-09 17:01:09 -08:00
z3fold.txt z3fold: the 3-fold allocator for compressed pages 2016-05-20 17:58:30 -07:00
zsmalloc.txt zsmalloc: zsmalloc documentation 2015-04-15 16:35:21 -07:00
zswap.txt zswap: update docs for runtime-changeable attributes 2015-09-10 13:29:01 -07:00