Patch series "Change the return value for page isolation functions", v3.
Now the page isolation functions did not return a boolean to indicate
success or not, instead it will return a negative error when failed
to isolate a page. So below code used in most places seem a boolean
success/failure thing, which can confuse people whether the isolation
is successful.
if (folio_isolate_lru(folio))
continue;
Moreover the page isolation functions only return 0 or -EBUSY, and
most users did not care about the negative error except for few users,
thus we can convert all page isolation functions to return a boolean
value, which can remove the confusion to make code more clear.
No functional changes intended in this patch series.
This patch (of 4):
Now the folio_isolate_lru() did not return a boolean value to indicate
isolation success or not, however below code checking the return value can
make people think that it was a boolean success/failure thing, which makes
people easy to make mistakes (see the fix patch[1]).
if (folio_isolate_lru(folio))
continue;
Thus it's better to check the negative error value expilictly returned by
folio_isolate_lru(), which makes code more clear per Linus's
suggestion[2]. Moreover Matthew suggested we can convert the isolation
functions to return a boolean[3], since most users did not care about the
negative error value, and can also remove the confusing of checking return
value.
So this patch converts the folio_isolate_lru() to return a boolean value,
which means return 'true' to indicate the folio isolation is successful,
and 'false' means a failure to isolation. Meanwhile changing all users'
logic of checking the isolation state.
No functional changes intended.
[1] https://lore.kernel.org/all/20230131063206.28820-1-Kuan-Ying.Lee@mediatek.com/T/#u
[2] https://lore.kernel.org/all/CAHk-=wiBrY+O-4=2mrbVyxR+hOqfdJ=Do6xoucfJ9_5az01L4Q@mail.gmail.com/
[3] https://lore.kernel.org/all/Y+sTFqwMNAjDvxw3@casper.infradead.org/
Link: https://lkml.kernel.org/r/cover.1676424378.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/8a4e3679ed4196168efadf7ea36c038f2f7d5aa9.1676424378.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: SeongJae Park <sj@kernel.org>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently there are two kmem-related helper functions with a confusing
semantics: memcg_kmem_enabled() and mem_cgroup_kmem_disabled().
The problem is that an obvious expectation
memcg_kmem_enabled() == !mem_cgroup_kmem_disabled(),
can be false.
mem_cgroup_kmem_disabled() is similar to mem_cgroup_disabled(): it returns
true only if CONFIG_MEMCG_KMEM is not set or the kmem accounting is
disabled using a boot time kernel option "cgroup.memory=nokmem". It never
changes the value dynamically.
memcg_kmem_enabled() is different: it always returns false until the first
non-root memory cgroup will get online (assuming the kernel memory
accounting is enabled). It's goal is to improve the performance on
systems without the cgroupfs mounted/memory controller enabled or on the
systems with only the root memory cgroup.
To make things more obvious and avoid potential bugs, let's rename
memcg_kmem_enabled() to memcg_kmem_online().
Link: https://lkml.kernel.org/r/20230213192922.1146370-1-roman.gushchin@linux.dev
Signed-off-by: Roman Gushchin <roman.gushchin@linux.dev>
Acked-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Dennis Zhou <dennis@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Recall that the per-node memcg LRU has two generations and they alternate
when the last memcg (of a given node) is moved from one to the other.
Each generation is also sharded into multiple bins to improve scalability.
A reclaimer starts with a random bin (in the old generation) and, if it
fails, it will retry, i.e., to try the rest of the bins.
If a reclaimer fails with the last memcg, it should move this memcg to the
young generation first, which causes the generations to alternate, and
then retry. Otherwise, the retries will be futile because all other bins
are empty.
Link: https://lkml.kernel.org/r/20230213075322.1416966-1-yuzhao@google.com
Fixes: e4dde56cd2 ("mm: multi-gen LRU: per-node lru_gen_folio lists")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: T.J. Mercier <tjmercier@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The debugfs_remove_recursive() is invoked by unregister_shrinker(), which
is holding the write lock of shrinker_rwsem. It will waits for the
handler of debugfs file complete. The handler also needs to hold the read
lock of shrinker_rwsem to do something. So it may cause the following
deadlock:
CPU0 CPU1
debugfs_file_get()
shrinker_debugfs_count_show()/shrinker_debugfs_scan_write()
unregister_shrinker()
--> down_write(&shrinker_rwsem);
debugfs_remove_recursive()
// wait for (A)
--> wait_for_completion();
// wait for (B)
--> down_read_killable(&shrinker_rwsem)
debugfs_file_put() -- (A)
up_write() -- (B)
The down_read_killable() can be killed, so that the above deadlock can be
recovered. But it still requires an extra kill action, otherwise it will
block all subsequent shrinker-related operations, so it's better to fix
it.
[akpm@linux-foundation.org: fix CONFIG_SHRINKER_DEBUG=n stub]
Link: https://lkml.kernel.org/r/20230202105612.64641-1-zhengqi.arch@bytedance.com
Fixes: 5035ebc644 ("mm: shrinkers: introduce debugfs interface for memory shrinkers")
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Reviewed-by: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Kent Overstreet <kent.overstreet@gmail.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Update the folio generation in place with or without
current->reclaim_state->mm_walk. The LRU lock is held for longer, if
mm_walk is NULL and the number of folios to update is more than
PAGEVEC_SIZE.
This causes a measurable regression from the LRU lock contention during a
microbencmark. But a tiny regression is not worth the complexity.
Link: https://lkml.kernel.org/r/20230118001827.1040870-8-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Improve readability of walk_pmd_range() and walk_pmd_range_locked().
Link: https://lkml.kernel.org/r/20230118001827.1040870-7-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Move memcg LRU code into a dedicated section. Improve the design doc to
outline its architecture.
Link: https://lkml.kernel.org/r/20230118001827.1040870-5-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Move Bloom filters code into a dedicated section. Improve the design doc
to explain Bloom filter usage and connection between aging and eviction in
their use.
Link: https://lkml.kernel.org/r/20230118001827.1040870-4-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add a section for lru_gen_look_around() in the code and the design doc.
Link: https://lkml.kernel.org/r/20230118001827.1040870-3-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm: multi-gen LRU: improve".
This patch series improves a few MGLRU functions, collects related
functions, and adds additional documentation.
This patch (of 7):
Add a section for working set protection in the code and the design doc.
The admin doc already contains its usage.
Link: https://lkml.kernel.org/r/20230118001827.1040870-1-talumbau@google.com
Link: https://lkml.kernel.org/r/20230118001827.1040870-2-talumbau@google.com
Signed-off-by: T.J. Alumbaugh <talumbau@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
lru_gen_migrate_mm() assumes lru_gen_add_mm() runs prior to itself. This
isn't true for the following scenario:
CPU 1 CPU 2
clone()
cgroup_can_fork()
cgroup_procs_write()
cgroup_post_fork()
task_lock()
lru_gen_migrate_mm()
task_unlock()
task_lock()
lru_gen_add_mm()
task_unlock()
And when the above happens, kernel crashes because of linked list
corruption (mm_struct->lru_gen.list).
Link: https://lore.kernel.org/r/20230115134651.30028-1-msizanoen@qtmlabs.xyz/
Link: https://lkml.kernel.org/r/20230116034405.2960276-1-yuzhao@google.com
Fixes: bd74fdaea1 ("mm: multi-gen LRU: support page table walks")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: msizanoen <msizanoen@qtmlabs.xyz>
Tested-by: msizanoen <msizanoen@qtmlabs.xyz>
Cc: <stable@vger.kernel.org> [6.1+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
This reverts commit 12a5d39552.
Although it is recognized that a finer grained pro-active reclaim is
something we need and want the semantic of this implementation is really
ambiguous.
In a follow up discussion it became clear that there are two essential
usecases here. One is to use memory.reclaim to pro-actively reclaim
memory and expectation is that the requested and reported amount of memory
is uncharged from the memcg. Another usecase focuses on pro-active
demotion when the memory is merely shuffled around to demotion targets
while the overall charged memory stays unchanged.
The current implementation considers demoted pages as reclaimed and that
break both usecases. [1] has tried to address the reporting part but
there are more issues with that summarized in [2] and follow up emails.
Let's revert the nodemask based extension of the memcg pro-active
reclaim for now until we settle with a more robust semantic.
[1] http://lkml.kernel.org/r/http://lkml.kernel.org/r/20221206023406.3182800-1-almasrymina@google.com
[2] http://lkml.kernel.org/r/Y5bsmpCyeryu3Zz1@dhcp22.suse.cz
Link: https://lkml.kernel.org/r/Y5xASNe1x8cusiTx@dhcp22.suse.cz
Fixes: 12a5d39552 ("mm: add nodes= arg to memory.reclaim")
Signed-off-by: Michal Hocko <mhocko@suse.com>
Cc: Bagas Sanjaya <bagasdotme@gmail.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Mina Almasry <almasrymina@google.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Wei Xu <weixugc@google.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Cc: zefan li <lizefan.x@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add vma_has_recency() to indicate whether a VMA may exhibit temporal
locality that the LRU algorithm relies on.
This function returns false for VMAs marked by VM_SEQ_READ or
VM_RAND_READ. While the former flag indicates linear access, i.e., a
special case of spatial locality, both flags indicate a lack of temporal
locality, i.e., the reuse of an area within a relatively small duration.
"Recency" is chosen over "locality" to avoid confusion between temporal
and spatial localities.
Before this patch, the active/inactive LRU only ignored the accessed bit
from VMAs marked by VM_SEQ_READ. After this patch, the active/inactive
LRU and MGLRU share the same logic: they both ignore the accessed bit if
vma_has_recency() returns false.
For the active/inactive LRU, the following fio test showed a [6, 8]%
increase in IOPS when randomly accessing mapped files under memory
pressure.
kb=$(awk '/MemTotal/ { print $2 }' /proc/meminfo)
kb=$((kb - 8*1024*1024))
modprobe brd rd_nr=1 rd_size=$kb
dd if=/dev/zero of=/dev/ram0 bs=1M
mkfs.ext4 /dev/ram0
mount /dev/ram0 /mnt/
swapoff -a
fio --name=test --directory=/mnt/ --ioengine=mmap --numjobs=8 \
--size=8G --rw=randrw --time_based --runtime=10m \
--group_reporting
The discussion that led to this patch is here [1]. Additional test
results are available in that thread.
[1] https://lore.kernel.org/r/Y31s%2FK8T85jh05wH@google.com/
Link: https://lkml.kernel.org/r/20221230215252.2628425-1-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Andrea Righi <andrea.righi@canonical.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Scanning page tables when hardware does not set the accessed bit has
no real use cases.
Link: https://lkml.kernel.org/r/20221222041905.2431096-9-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Among the flags in scan_control:
1. sc->may_swap, which indicates swap constraint due to memsw.max, is
supported as usual.
2. sc->proactive, which indicates reclaim by memory.reclaim, may not
opportunistically skip the aging path, since it is considered less
latency sensitive.
3. !(sc->gfp_mask & __GFP_IO), which indicates IO constraint, lowers
swappiness to prioritize file LRU, since clean file folios are more
likely to exist.
4. sc->may_writepage and sc->may_unmap, which indicates opportunistic
reclaim, are rejected, since unmapped clean folios are already
prioritized. Scanning for more of them is likely futile and can
cause high reclaim latency when there is a large number of memcgs.
The rest are handled by the existing code.
Link: https://lkml.kernel.org/r/20221222041905.2431096-8-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
For each node, memcgs are divided into two generations: the old and
the young. For each generation, memcgs are randomly sharded into
multiple bins to improve scalability. For each bin, an RCU hlist_nulls
is virtually divided into three segments: the head, the tail and the
default.
An onlining memcg is added to the tail of a random bin in the old
generation. The eviction starts at the head of a random bin in the old
generation. The per-node memcg generation counter, whose reminder (mod
2) indexes the old generation, is incremented when all its bins become
empty.
There are four operations:
1. MEMCG_LRU_HEAD, which moves an memcg to the head of a random bin in
its current generation (old or young) and updates its "seg" to
"head";
2. MEMCG_LRU_TAIL, which moves an memcg to the tail of a random bin in
its current generation (old or young) and updates its "seg" to
"tail";
3. MEMCG_LRU_OLD, which moves an memcg to the head of a random bin in
the old generation, updates its "gen" to "old" and resets its "seg"
to "default";
4. MEMCG_LRU_YOUNG, which moves an memcg to the tail of a random bin
in the young generation, updates its "gen" to "young" and resets
its "seg" to "default".
The events that trigger the above operations are:
1. Exceeding the soft limit, which triggers MEMCG_LRU_HEAD;
2. The first attempt to reclaim an memcg below low, which triggers
MEMCG_LRU_TAIL;
3. The first attempt to reclaim an memcg below reclaimable size
threshold, which triggers MEMCG_LRU_TAIL;
4. The second attempt to reclaim an memcg below reclaimable size
threshold, which triggers MEMCG_LRU_YOUNG;
5. Attempting to reclaim an memcg below min, which triggers
MEMCG_LRU_YOUNG;
6. Finishing the aging on the eviction path, which triggers
MEMCG_LRU_YOUNG;
7. Offlining an memcg, which triggers MEMCG_LRU_OLD.
Note that memcg LRU only applies to global reclaim, and the
round-robin incrementing of their max_seq counters ensures the
eventual fairness to all eligible memcgs. For memcg reclaim, it still
relies on mem_cgroup_iter().
Link: https://lkml.kernel.org/r/20221222041905.2431096-7-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Move should_run_aging() next to its only caller left.
Link: https://lkml.kernel.org/r/20221222041905.2431096-6-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Recall that the aging produces the youngest generation: first it scans
for accessed folios and updates their gen counters; then it increments
lrugen->max_seq.
The current aging fairness safeguard for kswapd uses two passes to
ensure the fairness to multiple eligible memcgs. On the first pass,
which is shared with the eviction, it checks whether all eligible
memcgs are low on cold folios. If so, it requires a second pass, on
which it ages all those memcgs at the same time.
With memcg LRU, the aging, while ensuring eventual fairness, will run
when necessary. Therefore the current aging fairness safeguard for
kswapd will not be needed.
Note that memcg LRU only applies to global reclaim. For memcg reclaim,
the aging can be unfair to different memcgs, i.e., their
lrugen->max_seq can be incremented at different paces.
Link: https://lkml.kernel.org/r/20221222041905.2431096-5-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Recall that the eviction consumes the oldest generation: first it
bucket-sorts folios whose gen counters were updated by the aging and
reclaims the rest; then it increments lrugen->min_seq.
The current eviction fairness safeguard for global reclaim has a
dilemma: when there are multiple eligible memcgs, should it continue
or stop upon meeting the reclaim goal? If it continues, it overshoots
and increases direct reclaim latency; if it stops, it loses fairness
between memcgs it has taken memory away from and those it has yet to.
With memcg LRU, the eviction, while ensuring eventual fairness, will
stop upon meeting its goal. Therefore the current eviction fairness
safeguard for global reclaim will not be needed.
Note that memcg LRU only applies to global reclaim. For memcg reclaim,
the eviction will continue, even if it is overshooting. This becomes
unconditional due to code simplification.
Link: https://lkml.kernel.org/r/20221222041905.2431096-4-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
lru_gen_folio will be chained into per-node lists by the coming
lrugen->list.
Link: https://lkml.kernel.org/r/20221222041905.2431096-3-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm: multi-gen LRU: memcg LRU", v3.
Overview
========
An memcg LRU is a per-node LRU of memcgs. It is also an LRU of LRUs,
since each node and memcg combination has an LRU of folios (see
mem_cgroup_lruvec()).
Its goal is to improve the scalability of global reclaim, which is
critical to system-wide memory overcommit in data centers. Note that
memcg reclaim is currently out of scope.
Its memory bloat is a pointer to each lruvec and negligible to each
pglist_data. In terms of traversing memcgs during global reclaim, it
improves the best-case complexity from O(n) to O(1) and does not affect
the worst-case complexity O(n). Therefore, on average, it has a sublinear
complexity in contrast to the current linear complexity.
The basic structure of an memcg LRU can be understood by an analogy to
the active/inactive LRU (of folios):
1. It has the young and the old (generations), i.e., the counterparts
to the active and the inactive;
2. The increment of max_seq triggers promotion, i.e., the counterpart
to activation;
3. Other events trigger similar operations, e.g., offlining an memcg
triggers demotion, i.e., the counterpart to deactivation.
In terms of global reclaim, it has two distinct features:
1. Sharding, which allows each thread to start at a random memcg (in
the old generation) and improves parallelism;
2. Eventual fairness, which allows direct reclaim to bail out at will
and reduces latency without affecting fairness over some time.
The commit message in patch 6 details the workflow:
https://lore.kernel.org/r/20221222041905.2431096-7-yuzhao@google.com/
The following is a simple test to quickly verify its effectiveness.
Test design:
1. Create multiple memcgs.
2. Each memcg contains a job (fio).
3. All jobs access the same amount of memory randomly.
4. The system does not experience global memory pressure.
5. Periodically write to the root memory.reclaim.
Desired outcome:
1. All memcgs have similar pgsteal counts, i.e., stddev(pgsteal)
over mean(pgsteal) is close to 0%.
2. The total pgsteal is close to the total requested through
memory.reclaim, i.e., sum(pgsteal) over sum(requested) is close
to 100%.
Actual outcome [1]:
MGLRU off MGLRU on
stddev(pgsteal) / mean(pgsteal) 75% 20%
sum(pgsteal) / sum(requested) 425% 95%
####################################################################
MEMCGS=128
for ((memcg = 0; memcg < $MEMCGS; memcg++)); do
mkdir /sys/fs/cgroup/memcg$memcg
done
start() {
echo $BASHPID > /sys/fs/cgroup/memcg$memcg/cgroup.procs
fio -name=memcg$memcg --numjobs=1 --ioengine=mmap \
--filename=/dev/zero --size=1920M --rw=randrw \
--rate=64m,64m --random_distribution=random \
--fadvise_hint=0 --time_based --runtime=10h \
--group_reporting --minimal
}
for ((memcg = 0; memcg < $MEMCGS; memcg++)); do
start &
done
sleep 600
for ((i = 0; i < 600; i++)); do
echo 256m >/sys/fs/cgroup/memory.reclaim
sleep 6
done
for ((memcg = 0; memcg < $MEMCGS; memcg++)); do
grep "pgsteal " /sys/fs/cgroup/memcg$memcg/memory.stat
done
####################################################################
[1]: This was obtained from running the above script (touches less
than 256GB memory) on an EPYC 7B13 with 512GB DRAM for over an
hour.
This patch (of 8):
The new name lru_gen_folio will be more distinct from the coming
lru_gen_memcg.
Link: https://lkml.kernel.org/r/20221222041905.2431096-1-yuzhao@google.com
Link: https://lkml.kernel.org/r/20221222041905.2431096-2-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michael Larabel <Michael@MichaelLarabel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Deactivate_page() has already been converted to use folios, this change
converts it to take in a folio argument instead of calling page_folio().
It also renames the function folio_deactivate() to be more consistent with
other folio functions.
[akpm@linux-foundation.org: fix left-over comments, per Yu Zhao]
Link: https://lkml.kernel.org/r/20221221180848.20774-5-vishal.moola@gmail.com
Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: SeongJae Park <sj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
* Randomize the per-cpu entry areas
Cleanups:
* Have CR3_ADDR_MASK use PHYSICAL_PAGE_MASK instead of open
coding it
* Move to "native" set_memory_rox() helper
* Clean up pmd_get_atomic() and i386-PAE
* Remove some unused page table size macros
-----BEGIN PGP SIGNATURE-----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=wwVF
-----END PGP SIGNATURE-----
Merge tag 'x86_mm_for_6.2_v2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 mm updates from Dave Hansen:
"New Feature:
- Randomize the per-cpu entry areas
Cleanups:
- Have CR3_ADDR_MASK use PHYSICAL_PAGE_MASK instead of open coding it
- Move to "native" set_memory_rox() helper
- Clean up pmd_get_atomic() and i386-PAE
- Remove some unused page table size macros"
* tag 'x86_mm_for_6.2_v2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (35 commits)
x86/mm: Ensure forced page table splitting
x86/kasan: Populate shadow for shared chunk of the CPU entry area
x86/kasan: Add helpers to align shadow addresses up and down
x86/kasan: Rename local CPU_ENTRY_AREA variables to shorten names
x86/mm: Populate KASAN shadow for entire per-CPU range of CPU entry area
x86/mm: Recompute physical address for every page of per-CPU CEA mapping
x86/mm: Rename __change_page_attr_set_clr(.checkalias)
x86/mm: Inhibit _PAGE_NX changes from cpa_process_alias()
x86/mm: Untangle __change_page_attr_set_clr(.checkalias)
x86/mm: Add a few comments
x86/mm: Fix CR3_ADDR_MASK
x86/mm: Remove P*D_PAGE_MASK and P*D_PAGE_SIZE macros
mm: Convert __HAVE_ARCH_P..P_GET to the new style
mm: Remove pointless barrier() after pmdp_get_lockless()
x86/mm/pae: Get rid of set_64bit()
x86_64: Remove pointless set_64bit() usage
x86/mm/pae: Be consistent with pXXp_get_and_clear()
x86/mm/pae: Use WRITE_ONCE()
x86/mm/pae: Don't (ab)use atomic64
mm/gup: Fix the lockless PMD access
...
I'd been worried by high "swapcached" counts in memcg OOM reports, thought
we had a problem freeing swapcache, but it was just the accounting that
was wrong.
Two issues:
1. When __remove_mapping() removes swapcache,
__delete_from_swap_cache() relies on memcg_data for the right counts to
be updated; but that had already been reset by mem_cgroup_swapout().
Swap those calls around - mem_cgroup_swapout() does not require the
swapcached flag to be set.
6.1 commit ac35a49023 ("mm: multi-gen LRU: minimal
implementation") already made a similar swap for workingset_eviction(),
but not for this.
2. memcg's "swapcached" count was added for memcg v2 stats, but
displayed on OOM even for memcg v1: so mem_cgroup_move_account() ought
to move it.
Link: https://lkml.kernel.org/r/b8b96ee0-1e1e-85f8-df97-c82a11d7cd14@google.com
Fixes: b603894248 ("mm: memcg: add swapcache stat for memcg v2")
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Shakeel Butt <shakeelb@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The nodes= arg instructs the kernel to only scan the given nodes for
proactive reclaim. For example use cases, consider a 2 tier memory
system:
nodes 0,1 -> top tier
nodes 2,3 -> second tier
$ echo "1m nodes=0" > memory.reclaim
This instructs the kernel to attempt to reclaim 1m memory from node 0.
Since node 0 is a top tier node, demotion will be attempted first. This
is useful to direct proactive reclaim to specific nodes that are under
pressure.
$ echo "1m nodes=2,3" > memory.reclaim
This instructs the kernel to attempt to reclaim 1m memory in the second
tier, since this tier of memory has no demotion targets the memory will be
reclaimed.
$ echo "1m nodes=0,1" > memory.reclaim
Instructs the kernel to reclaim memory from the top tier nodes, which can
be desirable according to the userspace policy if there is pressure on the
top tiers. Since these nodes have demotion targets, the kernel will
attempt demotion first.
Since commit 3f1509c57b ("Revert "mm/vmscan: never demote for memcg
reclaim""), the proactive reclaim interface memory.reclaim does both
reclaim and demotion. Reclaim and demotion incur different latency costs
to the jobs in the cgroup. Demoted memory would still be addressable by
the userspace at a higher latency, but reclaimed memory would need to
incur a pagefault.
The 'nodes' arg is useful to allow the userspace to control demotion and
reclaim independently according to its policy: if the memory.reclaim is
called on a node with demotion targets, it will attempt demotion first; if
it is called on a node without demotion targets, it will only attempt
reclaim.
Link: https://lkml.kernel.org/r/20221202223533.1785418-1-almasrymina@google.com
Signed-off-by: Mina Almasry <almasrymina@google.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Muchun Song <songmuchun@bytedance.com>
Cc: Bagas Sanjaya <bagasdotme@gmail.com>
Cc: "Huang, Ying" <ying.huang@intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Tejun Heo <tj@kernel.org>
Cc: Wei Xu <weixugc@google.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Cc: zefan li <lizefan.x@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reclaiming directly from top tier nodes breaks the aging pipeline of
memory tiers. If we have a RAM -> CXL -> storage hierarchy, we should
demote from RAM to CXL and from CXL to storage. If we reclaim a page from
RAM, it means we 'demote' it directly from RAM to storage, bypassing
potentially a huge amount of pages colder than it in CXL.
However disabling reclaim from top tier nodes entirely would cause ooms in
edge scenarios where lower tier memory is unreclaimable for whatever
reason, e.g. memory being mlocked() or too hot to reclaim. In these
cases we would rather the job run with a performance regression rather
than it oom altogether.
However, we can disable reclaim from top tier nodes for proactive reclaim.
That reclaim is not real memory pressure, and we don't have any cause to
be breaking the aging pipeline.
[akpm@linux-foundation.org: restore comment layout, per Ying Huang]
Link: https://lkml.kernel.org/r/20221201233317.1394958-1-almasrymina@google.com
Signed-off-by: Mina Almasry <almasrymina@google.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Wei Xu <weixugc@google.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm: memcg: fix protection of reclaim target memcg", v3.
This series fixes a bug in calculating the protection of the reclaim
target memcg where we end up using stale effective protection values from
the last reclaim operation, instead of completely ignoring the protection
of the reclaim target as intended. More detailed explanation and examples
in patch 1, which includes the fix. Patches 2 & 3 introduce a selftest
case that catches the bug.
This patch (of 3):
When we are doing memcg reclaim, the intended behavior is that we
ignore any protection (memory.min, memory.low) of the target memcg (but
not its children). Ever since the patch pointed to by the "Fixes" tag,
we actually read a stale value for the target memcg protection when
deciding whether to skip the memcg or not because it is protected. If
the stale value happens to be high enough, we don't reclaim from the
target memcg.
Essentially, in some cases we may falsely skip reclaiming from the
target memcg of reclaim because we read a stale protection value from
last time we reclaimed from it.
During reclaim, mem_cgroup_calculate_protection() is used to determine the
effective protection (emin and elow) values of a memcg. The protection of
the reclaim target is ignored, but we cannot set their effective
protection to 0 due to a limitation of the current implementation (see
comment in mem_cgroup_protection()). Instead, we leave their effective
protection values unchaged, and later ignore it in
mem_cgroup_protection().
However, mem_cgroup_protection() is called later in
shrink_lruvec()->get_scan_count(), which is after the
mem_cgroup_below_{min/low}() checks in shrink_node_memcgs(). As a result,
the stale effective protection values of the target memcg may lead us to
skip reclaiming from the target memcg entirely, before calling
shrink_lruvec(). This can be even worse with recursive protection, where
the stale target memcg protection can be higher than its standalone
protection. See two examples below (a similar version of example (a) is
added to test_memcontrol in a later patch).
(a) A simple example with proactive reclaim is as follows. Consider the
following hierarchy:
ROOT
|
A
|
B (memory.min = 10M)
Consider the following scenario:
- B has memory.current = 10M.
- The system undergoes global reclaim (or memcg reclaim in A).
- In shrink_node_memcgs():
- mem_cgroup_calculate_protection() calculates the effective min (emin)
of B as 10M.
- mem_cgroup_below_min() returns true for B, we do not reclaim from B.
- Now if we want to reclaim 5M from B using proactive reclaim
(memory.reclaim), we should be able to, as the protection of the
target memcg should be ignored.
- In shrink_node_memcgs():
- mem_cgroup_calculate_protection() immediately returns for B without
doing anything, as B is the target memcg, relying on
mem_cgroup_protection() to ignore B's stale effective min (still 10M).
- mem_cgroup_below_min() reads the stale effective min for B and we
skip it instead of ignoring its protection as intended, as we never
reach mem_cgroup_protection().
(b) An more complex example with recursive protection is as follows.
Consider the following hierarchy with memory_recursiveprot:
ROOT
|
A (memory.min = 50M)
|
B (memory.min = 10M, memory.high = 40M)
Consider the following scenario:
- B has memory.current = 35M.
- The system undergoes global reclaim (target memcg is NULL).
- B will have an effective min of 50M (all of A's unclaimed protection).
- B will not be reclaimed from.
- Now allocate 10M more memory in B, pushing it above it's high limit.
- The system undergoes memcg reclaim from B (target memcg is B).
- Like example (a), we do nothing in mem_cgroup_calculate_protection(),
then call mem_cgroup_below_min(), which will read the stale effective
min for B (50M) and skip it. In this case, it's even worse because we
are not just considering B's standalone protection (10M), but we are
reading a much higher stale protection (50M) which will cause us to not
reclaim from B at all.
This is an artifact of commit 45c7f7e1ef ("mm, memcg: decouple
e{low,min} state mutations from protection checks") which made
mem_cgroup_calculate_protection() only change the state without returning
any value. Before that commit, we used to return MEMCG_PROT_NONE for the
target memcg, which would cause us to skip the
mem_cgroup_below_{min/low}() checks. After that commit we do not return
anything and we end up checking the min & low effective protections for
the target memcg, which are stale.
Update mem_cgroup_supports_protection() to also check if we are reclaiming
from the target, and rename it to mem_cgroup_unprotected() (now returns
true if we should not protect the memcg, much simpler logic).
Link: https://lkml.kernel.org/r/20221202031512.1365483-1-yosryahmed@google.com
Link: https://lkml.kernel.org/r/20221202031512.1365483-2-yosryahmed@google.com
Fixes: 45c7f7e1ef ("mm, memcg: decouple e{low,min} state mutations from protection checks")
Signed-off-by: Yosry Ahmed <yosryahmed@google.com>
Reviewed-by: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Chris Down <chris@chrisdown.name>
Cc: David Rientjes <rientjes@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vasily Averin <vasily.averin@linux.dev>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Replace open-coded snprintf() with sysfs_emit() to simplify the code.
Link: https://lkml.kernel.org/r/202211241929015476424@zte.com.cn
Signed-off-by: Xu Panda <xu.panda@zte.com.cn>
Signed-off-by: Yang Yang <yang.yang29@zte.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
NODE_DATA() is preallocated for all possible nodes after commit
09f49dca57 ("mm: handle uninitialized numa nodes gracefully"). Checking
its return value against NULL is now unnecessary.
Link: https://lkml.kernel.org/r/20221116013808.3995280-2-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently, drop_caches are reclaiming node-by-node, looping on each node
until reclaim could not make progress. This can however leave quite some
slab entries (such as filesystem inodes) unreclaimed if objects say on
node 1 keep objects on node 0 pinned. So move the "loop until no
progress" loop to the node-by-node iteration to retry reclaim also on
other nodes if reclaim on some nodes made progress. This fixes problem
when drop_caches was not reclaiming lots of otherwise perfectly fine to
reclaim inodes.
Link: https://lkml.kernel.org/r/20221115123255.12559-1-jack@suse.cz
Signed-off-by: Jan Kara <jack@suse.cz>
Reported-by: You Zhou <you.zhou@intel.com>
Reported-by: Pengfei Xu <pengfei.xu@intel.com>
Tested-by: Pengfei Xu <pengfei.xu@intel.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Direct reclaim stats are useful for identifying a potential source for
application latency, as well as spotting issues with kswapd. However,
khugepaged currently distorts the picture: as a kernel thread it doesn't
impose allocation latencies on userspace, and it explicitly opts out of
kswapd reclaim. Its activity showing up in the direct reclaim stats is
misleading. Counting it as kswapd reclaim could also cause confusion when
trying to understand actual kswapd behavior.
Break out khugepaged from the direct reclaim counters into new
pgsteal_khugepaged, pgdemote_khugepaged, pgscan_khugepaged counters.
Test with a huge executable (CONFIG_READ_ONLY_THP_FOR_FS):
pgsteal_kswapd 1342185
pgsteal_direct 0
pgsteal_khugepaged 3623
pgscan_kswapd 1345025
pgscan_direct 0
pgscan_khugepaged 3623
Link: https://lkml.kernel.org/r/20221026180133.377671-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: Eric Bergen <ebergen@meta.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Yosry Ahmed <yosryahmed@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
When running as a Xen PV guests commit eed9a328aa ("mm: x86: add
CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG") can cause a protection violation in
pmdp_test_and_clear_young():
BUG: unable to handle page fault for address: ffff8880083374d0
#PF: supervisor write access in kernel mode
#PF: error_code(0x0003) - permissions violation
PGD 3026067 P4D 3026067 PUD 3027067 PMD 7fee5067 PTE 8010000008337065
Oops: 0003 [#1] PREEMPT SMP NOPTI
CPU: 7 PID: 158 Comm: kswapd0 Not tainted 6.1.0-rc5-20221118-doflr+ #1
RIP: e030:pmdp_test_and_clear_young+0x25/0x40
This happens because the Xen hypervisor can't emulate direct writes to
page table entries other than PTEs.
This can easily be fixed by introducing arch_has_hw_nonleaf_pmd_young()
similar to arch_has_hw_pte_young() and test that instead of
CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG.
Link: https://lkml.kernel.org/r/20221123064510.16225-1-jgross@suse.com
Fixes: eed9a328aa ("mm: x86: add CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG")
Signed-off-by: Juergen Gross <jgross@suse.com>
Reported-by: Sander Eikelenboom <linux@eikelenboom.it>
Acked-by: Yu Zhao <yuzhao@google.com>
Tested-by: Sander Eikelenboom <linux@eikelenboom.it>
Acked-by: David Hildenbrand <david@redhat.com> [core changes]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
balance_dirty_pages doesn't do the required dirty throttling on cgroupv1.
See commit 9badce000e ("cgroup, writeback: don't enable cgroup writeback
on traditional hierarchies"). Instead, the kernel depends on writeback
throttling in shrink_folio_list to achieve the same goal. With large
memory systems, the flusher may not be able to writeback quickly enough
such that we will start finding pages in the shrink_folio_list already in
writeback. Hence for cgroupv1 let's do a reclaim throttle after waking up
the flusher.
The below test which used to fail on a 256GB system completes till the the
file system is full with this change.
root@lp2:/sys/fs/cgroup/memory# mkdir test
root@lp2:/sys/fs/cgroup/memory# cd test/
root@lp2:/sys/fs/cgroup/memory/test# echo 120M > memory.limit_in_bytes
root@lp2:/sys/fs/cgroup/memory/test# echo $$ > tasks
root@lp2:/sys/fs/cgroup/memory/test# dd if=/dev/zero of=/home/kvaneesh/test bs=1M
Killed
Link: https://lkml.kernel.org/r/20221118070603.84081-1-aneesh.kumar@linux.ibm.com
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: zefan li <lizefan.x@bytedance.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The page reclaim isolates a batch of folios from the tail of one of the
LRU lists and works on those folios one by one. For a suitable
swap-backed folio, if the swap device is async, it queues that folio for
writeback. After the page reclaim finishes an entire batch, it puts back
the folios it queued for writeback to the head of the original LRU list.
In the meantime, the page writeback flushes the queued folios also by
batches. Its batching logic is independent from that of the page reclaim.
For each of the folios it writes back, the page writeback calls
folio_rotate_reclaimable() which tries to rotate a folio to the tail.
folio_rotate_reclaimable() only works for a folio after the page reclaim
has put it back. If an async swap device is fast enough, the page
writeback can finish with that folio while the page reclaim is still
working on the rest of the batch containing it. In this case, that folio
will remain at the head and the page reclaim will not retry it before
reaching there.
This patch adds a retry to evict_folios(). After evict_folios() has
finished an entire batch and before it puts back folios it cannot free
immediately, it retries those that may have missed the rotation.
Before this patch, ~60% of folios swapped to an Intel Optane missed
folio_rotate_reclaimable(). After this patch, ~99% of missed folios were
reclaimed upon retry.
This problem affects relatively slow async swap devices like Samsung 980
Pro much less and does not affect sync swap devices like zram or zswap at
all.
Link: https://lkml.kernel.org/r/20221116013808.3995280-1-yuzhao@google.com
Fixes: ac35a49023 ("mm: multi-gen LRU: minimal implementation")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: "Yin, Fengwei" <fengwei.yin@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
During proactive reclaim, we sometimes observe severe overreclaim, with
several thousand times more pages reclaimed than requested.
This trace was obtained from shrink_lruvec() during such an instance:
prio:0 anon_cost:1141521 file_cost:7767
nr_reclaimed:4387406 nr_to_reclaim:1047 (or_factor:4190)
nr=[7161123 345 578 1111]
While he reclaimer requested 4M, vmscan reclaimed close to 16G, most of it
by swapping. These requests take over a minute, during which the write()
to memory.reclaim is unkillably stuck inside the kernel.
Digging into the source, this is caused by the proportional reclaim
bailout logic. This code tries to resolve a fundamental conflict: to
reclaim roughly what was requested, while also aging all LRUs fairly and
in accordance to their size, swappiness, refault rates etc. The way it
attempts fairness is that once the reclaim goal has been reached, it stops
scanning the LRUs with the smaller remaining scan targets, and adjusts the
remainder of the bigger LRUs according to how much of the smaller LRUs was
scanned. It then finishes scanning that remainder regardless of the
reclaim goal.
This works fine if priority levels are low and the LRU lists are
comparable in size. However, in this instance, the cgroup that is
targeted by proactive reclaim has almost no files left - they've already
been squeezed out by proactive reclaim earlier - and the remaining anon
pages are hot. Anon rotations cause the priority level to drop to 0,
which results in reclaim targeting all of anon (a lot) and all of file
(almost nothing). By the time reclaim decides to bail, it has scanned
most or all of the file target, and therefor must also scan most or all of
the enormous anon target. This target is thousands of times larger than
the reclaim goal, thus causing the overreclaim.
The bailout code hasn't changed in years, why is this failing now? The
most likely explanations are two other recent changes in anon reclaim:
1. Before the series starting with commit 5df741963d ("mm: fix LRU
balancing effect of new transparent huge pages"), the VM was
overall relatively reluctant to swap at all, even if swap was
configured. This means the LRU balancing code didn't come into play
as often as it does now, and mostly in high pressure situations
where pronounced swap activity wouldn't be as surprising.
2. For historic reasons, shrink_lruvec() loops on the scan targets of
all LRU lists except the active anon one, meaning it would bail if
the only remaining pages to scan were active anon - even if there
were a lot of them.
Before the series starting with commit ccc5dc6734 ("mm/vmscan:
make active/inactive ratio as 1:1 for anon lru"), most anon pages
would live on the active LRU; the inactive one would contain only a
handful of preselected reclaim candidates. After the series, anon
gets aged similarly to file, and the inactive list is the default
for new anon pages as well, making it often the much bigger list.
As a result, the VM is now more likely to actually finish large
anon targets than before.
Change the code such that only one SWAP_CLUSTER_MAX-sized nudge toward the
larger LRU lists is made before bailing out on a met reclaim goal.
This fixes the extreme overreclaim problem.
Fairness is more subtle and harder to evaluate. No obvious misbehavior
was observed on the test workload, in any case. Conceptually, fairness
should primarily be a cumulative effect from regular, lower priority
scans. Once the VM is in trouble and needs to escalate scan targets to
make forward progress, fairness needs to take a backseat. This is also
acknowledged by the myriad exceptions in get_scan_count(). This patch
makes fairness decrease gradually, as it keeps fairness work static over
increasing priority levels with growing scan targets. This should make
more sense - although we may have to re-visit the exact values.
Link: https://lkml.kernel.org/r/20220802162811.39216-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Acked-by: Mel Gorman <mgorman@techsingularity.net>
Cc: Hugh Dickins <hughd@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
We noticed a 2% webserver throughput regression after upgrading from 5.6.
This could be tracked down to a shift in the anon/file reclaim balance
(confirmed with swappiness) that resulted in worse reclaim efficiency and
thus more kswapd activity for the same outcome.
The change that exposed the problem is aae466b005 ("mm/swap: implement
workingset detection for anonymous LRU"). By qualifying swapins based on
their refault distance, it lowered the cost of anon reclaim in this
workload, in turn causing (much) more anon scanning than before. Scanning
the anon list is more expensive due to the higher ratio of mmapped pages
that may rotate during reclaim, and so the result was an increase in %sys
time.
Right now, rotations aren't considered a cost when balancing scan pressure
between LRUs. We can end up with very few file refaults putting all the
scan pressure on hot anon pages that are rotated en masse, don't get
reclaimed, and never push back on the file LRU again. We still only
reclaim file cache in that case, but we burn a lot CPU rotating anon
pages. It's "fair" from an LRU age POV, but doesn't reflect the real cost
it imposes on the system.
Consider rotations as a secondary factor in balancing the LRUs. This
doesn't attempt to make a precise comparison between IO cost and CPU cost,
it just says: if reloads are about comparable between the lists, or
rotations are overwhelmingly different, adjust for CPU work.
This fixed the regression on our webservers. It has since been deployed
to the entire Meta fleet and hasn't caused any problems.
Link: https://lkml.kernel.org/r/20221013193113.726425-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Rik van Riel <riel@surriel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
wakeup_flusher_threads() was added under the assumption that if a system
runs out of clean cold pages, it might want to write back dirty pages more
aggressively so that they can become clean and be dropped.
However, doing so can breach the rate limit a system wants to impose on
writeback, resulting in early SSD wearout.
Link: https://lkml.kernel.org/r/YzSiWq9UEER5LKup@google.com
Fixes: bd74fdaea1 ("mm: multi-gen LRU: support page table walks")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: Axel Rasmussen <axelrasmussen@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
All callers now have a folio, so convert the function to take a folio.
Saves a couple of calls to compound_head().
Link: https://lkml.kernel.org/r/20220902194653.1739778-48-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add kernel-doc for folio_free_swap() and make it return bool. Add a
try_to_free_swap() compatibility wrapper.
Link: https://lkml.kernel.org/r/20220902194653.1739778-11-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "MM folio changes for 6.1", v2.
My focus this round has been on shmem. I believe it is now fully
converted to folios. Of course, shmem interacts with a lot of the swap
cache and other parts of the kernel, so there are patches all over the MM.
This patch series survives a round of xfstests on tmpfs, which is nice,
but hardly an exhaustive test. Hugh was nice enough to run a round of
tests on it and found a bug which is fixed in this edition.
This patch (of 57):
A lot of comments mention pages when they should say folios.
Fix them up.
[akpm@linux-foundation.org: fixups for mglru additions]
Link: https://lkml.kernel.org/r/20220902194653.1739778-1-willy@infradead.org
Link: https://lkml.kernel.org/r/20220902194653.1739778-2-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Currently, a higher tier node can only be demoted to selected nodes on the
next lower tier as defined by the demotion path. This strict demotion
order does not work in all use cases (e.g. some use cases may want to
allow cross-socket demotion to another node in the same demotion tier as a
fallback when the preferred demotion node is out of space). This demotion
order is also inconsistent with the page allocation fallback order when
all the nodes in a higher tier are out of space: The page allocation can
fall back to any node from any lower tier, whereas the demotion order
doesn't allow that currently.
This patch adds support to get all the allowed demotion targets for a
memory tier. demote_page_list() function is now modified to utilize this
allowed node mask as the fallback allocation mask.
Link: https://lkml.kernel.org/r/20220818131042.113280-9-aneesh.kumar@linux.ibm.com
Signed-off-by: Jagdish Gediya <jvgediya.oss@gmail.com>
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Acked-by: Wei Xu <weixugc@google.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Bharata B Rao <bharata@amd.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Hesham Almatary <hesham.almatary@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Tim Chen <tim.c.chen@intel.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: SeongJae Park <sj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
