Export try_to_free_mem_cgroup_pages function to allow vendor modules to reclaim a memory cgroup.
Bug: 192052083
Signed-off-by: Liujie Xie <xieliujie@oppo.com>
Change-Id: Iec6ef50f5c71c62d0c9aa6de90e56a143dac61c1
(cherry picked from commit a8385d61f2)
Some shrinker add lock in count_objects() may cause lock contention
issues and lead all task stall on slab shrink. Add vendor hook in
do_shrink_slab() for shrinker->count_objects() latency measuring.
Add 3 oem data in shrink_control struct. Two is for
shrinker->count_objects() and shrinker->scan_objects() latency
measuring, other one to store priority, some shrinker know the reclaimer
priority can control the memory reclaim more better.
Bug: 188684131
Change-Id: I80e9d90179bb52a99c54d9a067c6fcee835bb2ad
Signed-off-by: rongqianfeng <rongqianfeng@vivo.com>
Export shrink_slab to module for do shrink-memory action.
Bug: 221768451
Bug: 234405962
Signed-off-by: Peifeng Li <lipeifeng@oppo.com>
Change-Id: I5abe9ad419d64999b714d879c228625a243e90d1
Provide a vendor hook to allow page_referenced to be skipped
during shrink_active_list to avoid heavy cpuloading caused by
it.
Bug: 220878851
Bug: 234405962
Signed-off-by: Liujie Xie <xieliujie@oppo.com>
Signed-off-by: Peifeng Li <lipeifeng@oppo.com>
Change-Id: Ie0e369f8f8739fea59a95470af20ab0e976869d1
Add /sys/kernel/mm/lru_gen/enabled as a kill switch. Components that
can be disabled include:
0x0001: the multi-gen LRU core
0x0002: walking page table, when arch_has_hw_pte_young() returns
true
0x0004: clearing the accessed bit in non-leaf PMD entries, when
CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG=y
[yYnN]: apply to all the components above
E.g.,
echo y >/sys/kernel/mm/lru_gen/enabled
cat /sys/kernel/mm/lru_gen/enabled
0x0007
echo 5 >/sys/kernel/mm/lru_gen/enabled
cat /sys/kernel/mm/lru_gen/enabled
0x0005
NB: the page table walks happen on the scale of seconds under heavy
memory pressure, in which case the mmap_lock contention is a lesser
concern, compared with the LRU lock contention and the I/O congestion.
So far the only well-known case of the mmap_lock contention happens on
Android, due to Scudo [1] which allocates several thousand VMAs for
merely a few hundred MBs. The SPF and the Maple Tree also have
provided their own assessments [2][3]. However, if walking page tables
does worsen the mmap_lock contention, the kill switch can be used to
disable it. In this case the multi-gen LRU will suffer a minor
performance degradation, as shown previously.
Clearing the accessed bit in non-leaf PMD entries can also be
disabled, since this behavior was not tested on x86 varieties other
than Intel and AMD.
[1] https://source.android.com/devices/tech/debug/scudo
[2] https://lore.kernel.org/lkml/20220128131006.67712-1-michel@lespinasse.org/
[3] https://lore.kernel.org/lkml/20220202024137.2516438-1-Liam.Howlett@oracle.com/
Link: https://lore.kernel.org/lkml/20220309021230.721028-11-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I71801d9470a2588cad8bfd14fbcfafc7b010aa03
When multiple memcgs are available, it is possible to make better
choices based on generations and tiers and therefore improve the
overall performance under global memory pressure. This patch adds a
rudimentary optimization to select memcgs that can drop single-use
unmapped clean pages first. Doing so reduces the chance of going into
the aging path or swapping. These two operations can be costly.
A typical example that benefits from this optimization is a server
running mixed types of workloads, e.g., heavy anon workload in one
memcg and heavy buffered I/O workload in the other.
Though this optimization can be applied to both kswapd and direct
reclaim, it is only added to kswapd to keep the patchset manageable.
Later improvements will cover the direct reclaim path.
Server benchmark results:
Mixed workloads:
fio (buffered I/O): -[23, 25]%
IOPS BW
patch1-8: 2960k 11.3GiB/s
patch1-9: 2248k 8783MiB/s
memcached (anon): +[210, 214]%
Ops/sec KB/sec
patch1-8: 606940.09 23576.89
patch1-9: 1895197.49 73619.93
Mixed workloads:
fio (buffered I/O): -[4, 6]%
IOPS BW
5.18-ed4643521e6a: 2369k 9255MiB/s
patch1-9: 2248k 8783MiB/s
memcached (anon): +[510, 516]%
Ops/sec KB/sec
5.18-ed4643521e6a: 309189.58 12010.61
patch1-9: 1895197.49 73619.93
Configurations:
(changes since patch 6)
cat mixed.sh
modprobe brd rd_nr=2 rd_size=56623104
swapoff -a
mkswap /dev/ram0
swapon /dev/ram0
mkfs.ext4 /dev/ram1
mount -t ext4 /dev/ram1 /mnt
memtier_benchmark -S /var/run/memcached/memcached.sock \
-P memcache_binary -n allkeys --key-minimum=1 \
--key-maximum=50000000 --key-pattern=P:P -c 1 -t 36 \
--ratio 1:0 --pipeline 8 -d 2000
fio -name=mglru --numjobs=36 --directory=/mnt --size=1408m \
--buffered=1 --ioengine=io_uring --iodepth=128 \
--iodepth_batch_submit=32 --iodepth_batch_complete=32 \
--rw=randread --random_distribution=random --norandommap \
--time_based --ramp_time=10m --runtime=90m --group_reporting &
pid=$!
sleep 200
memtier_benchmark -S /var/run/memcached/memcached.sock \
-P memcache_binary -n allkeys --key-minimum=1 \
--key-maximum=50000000 --key-pattern=R:R -c 1 -t 36 \
--ratio 0:1 --pipeline 8 --randomize --distinct-client-seed
kill -INT $pid
wait
Client benchmark results:
no change (CONFIG_MEMCG=n)
Link: https://lore.kernel.org/lkml/20220309021230.721028-10-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I0641467dbd7c5ba0645602cec7fe8d6fdb750edb
To further exploit spatial locality, the aging prefers to walk page
tables to search for young PTEs and promote hot pages. A kill switch
will be added in the next patch to disable this behavior. When
disabled, the aging relies on the rmap only.
NB: this behavior has nothing similar with the page table scanning in
the 2.4 kernel [1], which searches page tables for old PTEs, adds cold
pages to swapcache and unmaps them.
To avoid confusion, the term "iteration" specifically means the
traversal of an entire mm_struct list; the term "walk" will be applied
to page tables and the rmap, as usual.
An mm_struct list is maintained for each memcg, and an mm_struct
follows its owner task to the new memcg when this task is migrated.
Given an lruvec, the aging iterates lruvec_memcg()->mm_list and calls
walk_page_range() with each mm_struct on this list to promote hot
pages before it increments max_seq.
When multiple page table walkers iterate the same list, each of them
gets a unique mm_struct; therefore they can run concurrently. Page
table walkers ignore any misplaced pages, e.g., if an mm_struct was
migrated, pages it left in the previous memcg will not be promoted
when its current memcg is under reclaim. Similarly, page table walkers
will not promote pages from nodes other than the one under reclaim.
This patch uses the following optimizations when walking page tables:
1. It tracks the usage of mm_struct's between context switches so that
page table walkers can skip processes that have been sleeping since
the last iteration.
2. It uses generational Bloom filters to record populated branches so
that page table walkers can reduce their search space based on the
query results, e.g., to skip page tables containing mostly holes or
misplaced pages.
3. It takes advantage of the accessed bit in non-leaf PMD entries when
CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG=y.
4. It does not zigzag between a PGD table and the same PMD table
spanning multiple VMAs. IOW, it finishes all the VMAs within the
range of the same PMD table before it returns to a PGD table. This
improves the cache performance for workloads that have large
numbers of tiny VMAs [2], especially when CONFIG_PGTABLE_LEVELS=5.
Server benchmark results:
Single workload:
fio (buffered I/O): no change
Single workload:
memcached (anon): +[5.5, 7.5]%
Ops/sec KB/sec
patch1-7: 1014393.57 39455.42
patch1-8: 1078507.59 41949.15
Configurations:
no change
Client benchmark results:
kswapd profiles:
patch1-7
45.54% lzo1x_1_do_compress (real work)
9.56% page_vma_mapped_walk
6.70% _raw_spin_unlock_irq
2.78% ptep_clear_flush
2.47% do_raw_spin_lock
2.22% __zram_bvec_write
1.87% lru_gen_look_around
1.78% memmove
1.77% obj_malloc
1.44% free_unref_page_list
patch1-8
47.02% lzo1x_1_do_compress (real work)
6.73% page_vma_mapped_walk
6.14% _raw_spin_unlock_irq
3.39% walk_pte_range
2.63% ptep_clear_flush
2.29% __zram_bvec_write
2.10% do_raw_spin_lock
1.81% memmove
1.73% obj_malloc
1.53% free_unref_page_list
Configurations:
no change
[1] https://lwn.net/Articles/23732/
[2] https://source.android.com/devices/tech/debug/scudo
Link: https://lore.kernel.org/lkml/20220309021230.721028-9-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I5a3c97cf8ebf8d65d5f9528cd979a637c190053e
Searching the rmap for PTEs mapping each page on an LRU list (to test
and clear the accessed bit) can be expensive because pages from
different VMAs (PA space) are not cache friendly to the rmap (VA
space). For workloads mostly using mapped pages, the rmap has a high
CPU cost in the reclaim path.
This patch exploits spatial locality to reduce the trips into the
rmap. When shrink_page_list() walks the rmap and finds a young PTE, a
new function lru_gen_look_around() scans at most BITS_PER_LONG-1
adjacent PTEs. On finding another young PTE, it clears the accessed
bit and updates the gen counter of the page mapped by this PTE to
(max_seq%MAX_NR_GENS)+1.
Server benchmark results:
Single workload:
fio (buffered I/O): no change
Single workload:
memcached (anon): +[4, 6]%
Ops/sec KB/sec
patch1-6: 964656.80 37520.88
patch1-7: 1014393.57 39455.42
Configurations:
no change
Client benchmark results:
kswapd profiles:
patch1-6
36.13% lzo1x_1_do_compress (real work)
19.16% page_vma_mapped_walk
6.55% _raw_spin_unlock_irq
4.02% do_raw_spin_lock
2.32% anon_vma_interval_tree_iter_first
2.11% ptep_clear_flush
1.76% __zram_bvec_write
1.64% folio_referenced_one
1.40% memmove
1.35% obj_malloc
patch1-7
45.54% lzo1x_1_do_compress (real work)
9.56% page_vma_mapped_walk
6.70% _raw_spin_unlock_irq
2.78% ptep_clear_flush
2.47% do_raw_spin_lock
2.22% __zram_bvec_write
1.87% lru_gen_look_around
1.78% memmove
1.77% obj_malloc
1.44% free_unref_page_list
Configurations:
no change
Link: https://lore.kernel.org/lkml/20220309021230.721028-8-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I9a290343840f3cf925c891c8e360c7cdc24ffb9c
To avoid confusion, the terms "promotion" and "demotion" will be
applied to the multi-gen LRU, as a new convention; the terms
"activation" and "deactivation" will be applied to the active/inactive
LRU, as usual.
The aging produces young generations. Given an lruvec, it increments
max_seq when max_seq-min_seq+1 approaches MIN_NR_GENS. The aging
promotes hot pages to the youngest generation when it finds them
accessed through page tables; the demotion of cold pages happens
consequently when it increments max_seq. The aging has the complexity
O(nr_hot_pages), since it is only interested in hot pages. Promotion
in the aging path does not require any LRU list operations, only the
updates of the gen counter and lrugen->nr_pages[]; demotion, unless as
the result of the increment of max_seq, requires LRU list operations,
e.g., lru_deactivate_fn().
The eviction consumes old generations. Given an lruvec, it increments
min_seq when the lists indexed by min_seq%MAX_NR_GENS become empty. A
feedback loop modeled after the PID controller monitors refaults over
anon and file types and decides which type to evict when both types
are available from the same generation.
Each generation is divided into multiple tiers. Tiers represent
different ranges of numbers of accesses through file descriptors. A
page accessed N times through file descriptors is in tier
order_base_2(N). Tiers do not have dedicated lrugen->lists[], only
bits in page->flags. In contrast to moving across generations, which
requires the LRU lock, moving across tiers only involves operations on
page->flags. The feedback loop also monitors refaults over all tiers
and decides when to protect pages in which tiers (N>1), using the
first tier (N=0,1) as a baseline. The first tier contains single-use
unmapped clean pages, which are most likely the best choices. The
eviction moves a page to the next generation, i.e., min_seq+1, if the
feedback loop decides so. This approach has the following advantages:
1. It removes the cost of activation in the buffered access path by
inferring whether pages accessed multiple times through file
descriptors are statistically hot and thus worth protecting in the
eviction path.
2. It takes pages accessed through page tables into account and avoids
overprotecting pages accessed multiple times through file
descriptors. (Pages accessed through page tables are in the first
tier, since N=0.)
3. More tiers provide better protection for pages accessed more than
twice through file descriptors, when under heavy buffered I/O
workloads.
Server benchmark results:
Single workload:
fio (buffered I/O): +[38, 40]%
IOPS BW
5.18-ed4643521e6a: 2547k 9989MiB/s
patch1-6: 3540k 13.5GiB/s
Single workload:
memcached (anon): +[103, 107]%
Ops/sec KB/sec
5.18-ed4643521e6a: 469048.66 18243.91
patch1-6: 964656.80 37520.88
Configurations:
CPU: two Xeon 6154
Mem: total 256G
Node 1 was only used as a ram disk to reduce the variance in the
results.
patch drivers/block/brd.c <<EOF
99,100c99,100
< gfp_flags = GFP_NOIO | __GFP_ZERO | __GFP_HIGHMEM;
< page = alloc_page(gfp_flags);
---
> gfp_flags = GFP_NOIO | __GFP_ZERO | __GFP_HIGHMEM | __GFP_THISNODE;
> page = alloc_pages_node(1, gfp_flags, 0);
EOF
cat >>/etc/systemd/system.conf <<EOF
CPUAffinity=numa
NUMAPolicy=bind
NUMAMask=0
EOF
cat >>/etc/memcached.conf <<EOF
-m 184320
-s /var/run/memcached/memcached.sock
-a 0766
-t 36
-B binary
EOF
cat fio.sh
modprobe brd rd_nr=1 rd_size=113246208
swapoff -a
mkfs.ext4 /dev/ram0
mount -t ext4 /dev/ram0 /mnt
mkdir /sys/fs/cgroup/user.slice/test
echo 38654705664 >/sys/fs/cgroup/user.slice/test/memory.max
echo $$ >/sys/fs/cgroup/user.slice/test/cgroup.procs
fio -name=mglru --numjobs=72 --directory=/mnt --size=1408m \
--buffered=1 --ioengine=io_uring --iodepth=128 \
--iodepth_batch_submit=32 --iodepth_batch_complete=32 \
--rw=randread --random_distribution=random --norandommap \
--time_based --ramp_time=10m --runtime=5m --group_reporting
cat memcached.sh
modprobe brd rd_nr=1 rd_size=113246208
swapoff -a
mkswap /dev/ram0
swapon /dev/ram0
memtier_benchmark -S /var/run/memcached/memcached.sock \
-P memcache_binary -n allkeys --key-minimum=1 \
--key-maximum=65000000 --key-pattern=P:P -c 1 -t 36 \
--ratio 1:0 --pipeline 8 -d 2000
memtier_benchmark -S /var/run/memcached/memcached.sock \
-P memcache_binary -n allkeys --key-minimum=1 \
--key-maximum=65000000 --key-pattern=R:R -c 1 -t 36 \
--ratio 0:1 --pipeline 8 --randomize --distinct-client-seed
Client benchmark results:
kswapd profiles:
5.18-ed4643521e6a
39.56% page_vma_mapped_walk
19.32% lzo1x_1_do_compress (real work)
7.18% do_raw_spin_lock
4.23% _raw_spin_unlock_irq
2.26% vma_interval_tree_subtree_search
2.12% vma_interval_tree_iter_next
2.11% folio_referenced_one
1.90% anon_vma_interval_tree_iter_first
1.47% ptep_clear_flush
0.97% __anon_vma_interval_tree_subtree_search
patch1-6
36.13% lzo1x_1_do_compress (real work)
19.16% page_vma_mapped_walk
6.55% _raw_spin_unlock_irq
4.02% do_raw_spin_lock
2.32% anon_vma_interval_tree_iter_first
2.11% ptep_clear_flush
1.76% __zram_bvec_write
1.64% folio_referenced_one
1.40% memmove
1.35% obj_malloc
Configurations:
CPU: single Snapdragon 7c
Mem: total 4G
Chrome OS MemoryPressure [1]
[1] https://chromium.googlesource.com/chromiumos/platform/tast-tests/
Link: https://lore.kernel.org/lkml/20220309021230.721028-7-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I3fe4850006d7984cd9f4fd46134b826609dc2f86
Evictable pages are divided into multiple generations for each lruvec.
The youngest generation number is stored in lrugen->max_seq for both
anon and file types as they are aged on an equal footing. The oldest
generation numbers are stored in lrugen->min_seq[] separately for anon
and file types as clean file pages can be evicted regardless of swap
constraints. These three variables are monotonically increasing.
Generation numbers are truncated into order_base_2(MAX_NR_GENS+1) bits
in order to fit into the gen counter in page->flags. Each truncated
generation number is an index to lrugen->lists[]. The sliding window
technique is used to track at least MIN_NR_GENS and at most
MAX_NR_GENS generations. The gen counter stores a value within [1,
MAX_NR_GENS] while a page is on one of lrugen->lists[]. Otherwise it
stores 0.
There are two conceptually independent procedures: "the aging", which
produces young generations, and "the eviction", which consumes old
generations. They form a closed-loop system, i.e., "the page reclaim".
Both procedures can be invoked from userspace for the purposes of
working set estimation and proactive reclaim. These features are
required to optimize job scheduling (bin packing) in data centers. The
variable size of the sliding window is designed for such use cases
[1][2].
To avoid confusion, the terms "hot" and "cold" will be applied to the
multi-gen LRU, as a new convention; the terms "active" and "inactive"
will be applied to the active/inactive LRU, as usual.
The protection of hot pages and the selection of cold pages are based
on page access channels and patterns. There are two access channels:
one through page tables and the other through file descriptors. The
protection of the former channel is by design stronger because:
1. The uncertainty in determining the access patterns of the former
channel is higher due to the approximation of the accessed bit.
2. The cost of evicting the former channel is higher due to the TLB
flushes required and the likelihood of encountering the dirty bit.
3. The penalty of underprotecting the former channel is higher because
applications usually do not prepare themselves for major page
faults like they do for blocked I/O. E.g., GUI applications
commonly use dedicated I/O threads to avoid blocking the rendering
threads.
There are also two access patterns: one with temporal locality and the
other without. For the reasons listed above, the former channel is
assumed to follow the former pattern unless VM_SEQ_READ or
VM_RAND_READ is present; the latter channel is assumed to follow the
latter pattern unless outlying refaults have been observed [3][4].
The next patch will address the "outlying refaults". Three macros,
i.e., LRU_REFS_WIDTH, LRU_REFS_PGOFF and LRU_REFS_MASK, used later are
added in this patch to make the entire patchset less diffy.
A page is added to the youngest generation on faulting. The aging
needs to check the accessed bit at least twice before handing this
page over to the eviction. The first check takes care of the accessed
bit set on the initial fault; the second check makes sure this page
has not been used since then. This protocol, AKA second chance,
requires a minimum of two generations, hence MIN_NR_GENS.
[1] https://dl.acm.org/doi/10.1145/3297858.3304053
[2] https://dl.acm.org/doi/10.1145/3503222.3507731
[3] https://lwn.net/Articles/495543/
[4] https://lwn.net/Articles/815342/
Link: https://lore.kernel.org/lkml/20220309021230.721028-6-yuzhao@google.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Acked-by: Brian Geffon <bgeffon@google.com>
Acked-by: Jan Alexander Steffens (heftig) <heftig@archlinux.org>
Acked-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Acked-by: Steven Barrett <steven@liquorix.net>
Acked-by: Suleiman Souhlal <suleiman@google.com>
Tested-by: Daniel Byrne <djbyrne@mtu.edu>
Tested-by: Donald Carr <d@chaos-reins.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: Konstantin Kharlamov <Hi-Angel@yandex.ru>
Tested-by: Shuang Zhai <szhai2@cs.rochester.edu>
Tested-by: Sofia Trinh <sofia.trinh@edi.works>
Tested-by: Vaibhav Jain <vaibhav@linux.ibm.com>
Bug: 227651406
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Change-Id: I333ec6a1d2abfa60d93d6adc190ed3eefe441512
Static code analysis tool reported NULL pointer access in
shrink_page_list() as the commit 26aa2d199d ("mm/migrate: demote
pages during reclaim") expects valid pgdat.
There is already an existing api, reclaim_pages, that tries to reclaim
pages from the list. use it instead of creating custom function.
Bug: 201263305
Fixes: 96f80f6284 ("ANDROID: mm: add reclaim_shmem_address_space() for faster reclaims")
Change-Id: Iaa11feac94c9e8338324ace0276c49d6a0adeb0c
Signed-off-by: Charan Teja Reddy <quic_charante@quicinc.com>
Page replacement is handled in the Linux Kernel in one of two ways:
1) Asynchronously via kswapd
2) Synchronously, via direct reclaim
At page allocation time the allocating task is immediately given a page
from the zone free list allowing it to go right back to work doing
whatever it was doing; Probably directly or indirectly executing business
logic.
Just prior to satisfying the allocation, free pages is checked to see if
it has reached the zone low watermark and if so, kswapd is awakened.
Kswapd will start scanning pages looking for inactive pages to evict to
make room for new page allocations. The work of kswapd allows tasks to
continue allocating memory from their respective zone free list without
incurring any delay.
When the demand for free pages exceeds the rate that kswapd tasks can
supply them, page allocation works differently. Once the allocating task
finds that the number of free pages is at or below the zone min watermark,
the task will no longer pull pages from the free list. Instead, the task
will run the same CPU-bound routines as kswapd to satisfy its own
allocation by scanning and evicting pages. This is called a direct reclaim.
The time spent performing a direct reclaim can be substantial, often
taking tens to hundreds of milliseconds for small order0 allocations to
half a second or more for order9 huge-page allocations. In fact, kswapd is
not actually required on a linux system. It exists for the sole purpose of
optimizing performance by preventing direct reclaims.
When memory shortfall is sufficient to trigger direct reclaims, they can
occur in any task that is running on the system. A single aggressive
memory allocating task can set the stage for collateral damage to occur in
small tasks that rarely allocate additional memory. Consider the impact of
injecting an additional 100ms of latency when nscd allocates memory to
facilitate caching of a DNS query.
The presence of direct reclaims 10 years ago was a fairly reliable
indicator that too much was being asked of a Linux system. Kswapd was
likely wasting time scanning pages that were ineligible for eviction.
Adding RAM or reducing the working set size would usually make the problem
go away. Since then hardware has evolved to bring a new struggle for
kswapd. Storage speeds have increased by orders of magnitude while CPU
clock speeds stayed the same or even slowed down in exchange for more
cores per package. This presents a throughput problem for a single
threaded kswapd that will get worse with each generation of new hardware.
Test Details
NOTE: The tests below were run with shadow entries disabled. See the
associated patch and cover letter for details
The tests below were designed with the assumption that a kswapd bottleneck
is best demonstrated using filesystem reads. This way, the inactive list
will be full of clean pages, simplifying the analysis and allowing kswapd
to achieve the highest possible steal rate. Maximum steal rates for kswapd
are likely to be the same or lower for any other mix of page types on the
system.
Tests were run on a 2U Oracle X7-2L with 52 Intel Xeon Skylake 2GHz cores,
756GB of RAM and 8 x 3.6 TB NVMe Solid State Disk drives. Each drive has
an XFS file system mounted separately as /d0 through /d7. SSD drives
require multiple concurrent streams to show their potential, so I created
eleven 250GB zero-filled files on each drive so that I could test with
parallel reads.
The test script runs in multiple stages. At each stage, the number of dd
tasks run concurrently is increased by 2. I did not include all of the
test output for brevity.
During each stage dd tasks are launched to read from each drive in a round
robin fashion until the specified number of tasks for the stage has been
reached. Then iostat, vmstat and top are started in the background with 10
second intervals. After five minutes, all of the dd tasks are killed and
the iostat, vmstat and top output is parsed in order to report the
following:
CPU consumption
- sy - aggregate kernel mode CPU consumption from vmstat output. The value
doesn't tend to fluctuate much so I just grab the highest value.
Each sample is averaged over 10 seconds
- dd_cpu - for all of the dd tasks averaged across the top samples since
there is a lot of variation.
Throughput
- in Kbytes
- Command is iostat -x -d 10 -g total
This first test performs reads using O_DIRECT in order to show the maximum
throughput that can be obtained using these drives. It also demonstrates
how rapidly throughput scales as the number of dd tasks are increased.
The dd command for this test looks like this:
Command Used: dd iflag=direct if=/d${i}/$n of=/dev/null bs=4M
Test #1: Direct IO
dd sy dd_cpu throughput
6 0 2.33 14726026.40
10 1 2.95 19954974.80
16 1 2.63 24419689.30
22 1 2.63 25430303.20
28 1 2.91 26026513.20
34 1 2.53 26178618.00
40 1 2.18 26239229.20
46 1 1.91 26250550.40
52 1 1.69 26251845.60
58 1 1.54 26253205.60
64 1 1.43 26253780.80
70 1 1.31 26254154.80
76 1 1.21 26253660.80
82 1 1.12 26254214.80
88 1 1.07 26253770.00
90 1 1.04 26252406.40
Throughput was close to peak with only 22 dd tasks. Very little system CPU
was consumed as expected as the drives DMA directly into the user address
space when using direct IO.
In this next test, the iflag=direct option is removed and we only run the
test until the pgscan_kswapd from /proc/vmstat starts to increment. At
that point metrics are parsed and reported and the pagecache contents are
dropped prior to the next test. Lather, rinse, repeat.
Test #2: standard file system IO, no page replacement
dd sy dd_cpu throughput
6 2 28.78 5134316.40
10 3 31.40 8051218.40
16 5 34.73 11438106.80
22 7 33.65 14140596.40
28 8 31.24 16393455.20
34 10 29.88 18219463.60
40 11 28.33 19644159.60
46 11 25.05 20802497.60
52 13 26.92 22092370.00
58 13 23.29 22884881.20
64 14 23.12 23452248.80
70 15 22.40 23916468.00
76 16 22.06 24328737.20
82 17 20.97 24718693.20
88 16 18.57 25149404.40
90 16 18.31 25245565.60
Each read has to pause after the buffer in kernel space is populated while
those pages are added to the pagecache and copied into the user address
space. For this reason, more parallel streams are required to achieve peak
throughput. The copy operation consumes substantially more CPU than direct
IO as expected.
The next test measures throughput after kswapd starts running. This is the
same test only we wait for kswapd to wake up before we start collecting
metrics. The script actually keeps track of a few things that were not
mentioned earlier. It tracks direct reclaims and page scans by watching
the metrics in /proc/vmstat. CPU consumption for kswapd is tracked the
same way it is tracked for dd.
Since the test is 100% reads, you can assume that the page steal rate for
kswapd and direct reclaims is almost identical to the scan rate.
Test #3: 1 kswapd thread per node
dd sy dd_cpu kswapd0 kswapd1 throughput dr pgscan_kswapd pgscan_direct
10 4 26.07 28.56 27.03 7355924.40 0 459316976 0
16 7 34.94 69.33 69.66 10867895.20 0 872661643 0
22 10 36.03 93.99 99.33 13130613.60 489 1037654473 11268334
28 10 30.34 95.90 98.60 14601509.60 671 1182591373 15429142
34 14 34.77 97.50 99.23 16468012.00 10850 1069005644 249839515
40 17 36.32 91.49 97.11 17335987.60 18903 975417728 434467710
46 19 38.40 90.54 91.61 17705394.40 25369 855737040 582427973
52 22 40.88 83.97 83.70 17607680.40 31250 709532935 724282458
58 25 40.89 82.19 80.14 17976905.60 35060 657796473 804117540
64 28 41.77 73.49 75.20 18001910.00 39073 561813658 895289337
70 33 45.51 63.78 64.39 17061897.20 44523 379465571 1020726436
76 36 46.95 57.96 60.32 16964459.60 47717 291299464 1093172384
82 39 47.16 55.43 56.16 16949956.00 49479 247071062 1134163008
88 42 47.41 53.75 47.62 16930911.20 51521 195449924 1180442208
90 43 47.18 51.40 50.59 16864428.00 51618 190758156 1183203901
In the previous test where kswapd was not involved, the system-wide kernel
mode CPU consumption with 90 dd tasks was 16%. In this test CPU consumption
with 90 tasks is at 43%. With 52 cores, and two kswapd tasks (one per NUMA
node), kswapd can only be responsible for a little over 4% of the increase.
The rest is likely caused by 51,618 direct reclaims that scanned 1.2
billion pages over the five minute time period of the test.
Same test, more kswapd tasks:
Test #4: 4 kswapd threads per node
dd sy dd_cpu kswapd0 kswapd1 throughput dr pgscan_kswapd pgscan_direct
10 5 27.09 16.65 14.17 7842605.60 0 459105291 0
16 10 37.12 26.02 24.85 11352920.40 15 920527796 358515
22 11 36.94 37.13 35.82 13771869.60 0 1132169011 0
28 13 35.23 48.43 46.86 16089746.00 0 1312902070 0
34 15 33.37 53.02 55.69 18314856.40 0 1476169080 0
40 19 35.90 69.60 64.41 19836126.80 0 1629999149 0
46 22 36.82 88.55 57.20 20740216.40 0 1708478106 0
52 24 34.38 93.76 68.34 21758352.00 0 1794055559 0
58 24 30.51 79.20 82.33 22735594.00 0 1872794397 0
64 26 30.21 97.12 76.73 23302203.60 176 1916593721 4206821
70 33 32.92 92.91 92.87 23776588.00 3575 1817685086 85574159
76 37 31.62 91.20 89.83 24308196.80 4752 1812262569 113981763
82 29 25.53 93.23 92.33 24802791.20 306 2032093122 7350704
88 43 37.12 76.18 77.01 25145694.40 20310 1253204719 487048202
90 42 38.56 73.90 74.57 22516787.60 22774 1193637495 545463615
By increasing the number of kswapd threads, throughput increased by ~50%
while kernel mode CPU utilization decreased or stayed the same, likely due
to a decrease in the number of parallel tasks at any given time doing page
replacement.
Signed-off-by: Buddy Lumpkin <buddy.lumpkin@oracle.com>
Bug: 201263306
Link: https://lore.kernel.org/lkml/1522661062-39745-1-git-send-email-buddy.lumpkin@oracle.com
[charante@codeaurora.org]: Changes made to select number of kswapds through uapi
Change-Id: I8425cab7f40cbeaf65af0ea118c1a9ac7da0930e
Signed-off-by: Charan Teja Reddy <charante@codeaurora.org>
[quic_vjitta@quicinc.com]: Changes made to move multiple kswapd threads logic to vendor hooks
Signed-off-by: Vijayanand Jitta <quic_vjitta@quicinc.com>
(cherry picked from commit 0d61a651e4)
To support multiple kswap threads vendor modules need
access to kswapd function. So, export it.
Bug: 201263306
Change-Id: I442612710835f39836a295e9d1936f86826ab960
Signed-off-by: Vijayanand Jitta <quic_vjitta@quicinc.com>
Add the functionality that allow users of shmem to reclaim its pages
without going through the kswapd/direct reclaim path. An example usecase
is: Say that device allocates a larger amount of shmem pages and shares
it with hardware. To faster reclaims such pages, drivers can register
the shrinkers and call reclaim_shmem_address_space().
The implementation of this function is mostly borrowed from
reclaim_address_space() implemented for per process reclaim[1].
[1] https://lore.kernel.org/patchwork/cover/378056/
Bug: 201263305
Change-Id: I03d2c3b9610612af977f89ddeabb63b8e9e50918
Signed-off-by: Charan Teja Reddy <quic_charante@quicinc.com>
When performing memory reclaim, support treating anonymous and
file backed pages equally. Swapping anonymous pages out to memory
can be efficient enough to justify treating anonymous and file backed
pages equally.
Create a vendor hook inside of get_scan_count so that equal reclaim of
anon and file pages can be enabled inside of the trace hook based on
certain conditions.
Bug: 175415908
Change-Id: Idac2f1468371549d20dd3e5652c7382dc3d7d9cf
Signed-off-by: Sudarshan Rajagopalan <sudaraja@codeaurora.org>
Signed-off-by: Isaac J. Manjarres <isaacm@codeaurora.org>
Commit f56ce412a5 ("mm: memcontrol: fix occasional OOMs due to
proportional memory.low reclaim") introduced a divide by zero corner
case when oomd is being used in combination with cgroup memory.low
protection.
When oomd decides to kill a cgroup, it will force the cgroup memory to
be reclaimed after killing the tasks, by writing to the memory.max file
for that cgroup, forcing the remaining page cache and reclaimable slab
to be reclaimed down to zero.
Previously, on cgroups with some memory.low protection that would result
in the memory being reclaimed down to the memory.low limit, or likely
not at all, having the page cache reclaimed asynchronously later.
With f56ce412a5 the oomd write to memory.max tries to reclaim all the
way down to zero, which may race with another reclaimer, to the point of
ending up with the divide by zero below.
This patch implements the obvious fix.
Link: https://lkml.kernel.org/r/20210826220149.058089c6@imladris.surriel.com
Fixes: f56ce412a5 ("mm: memcontrol: fix occasional OOMs due to proportional memory.low reclaim")
Signed-off-by: Rik van Riel <riel@surriel.com>
Acked-by: Roman Gushchin <guro@fb.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Chris Down <chris@chrisdown.name>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
drop_slab_node() is called as part of echo 2>/proc/sys/vm/drop_caches
operation. It iterates over all memcgs and calls shrink_slab() which in
turn iterates over all slab shrinkers. Freed objects are counted and as
long as the total number of freed objects from all memcgs and shrinkers is
higher than 10, drop_slab_node() loops for another full memcgs*shrinkers
iteration.
This arbitrary constant threshold of 10 can result in effectively an
infinite loop on a system with large number of memcgs and/or parallel
activity that allocates new objects. This has been reported previously by
Chunxin Zang [1] and recently by our customer.
The previous report [1] has resulted in commit 069c411de4 ("mm/vmscan:
fix infinite loop in drop_slab_node") which added a check for signals
allowing the user to terminate the command writing to drop_caches. At the
time it was also considered to make the threshold grow with each iteration
to guarantee termination, but such patch hasn't been formally proposed
yet.
This patch implements the dynamically growing threshold. At first
iteration it's enough to free one object to continue, and this threshold
effectively doubles with each iteration. Our customer's feedback was
positive.
There is always a risk that this change will result on some system in a
previously terminating drop_caches operation to terminate sooner and free
fewer objects. Ideally the semantics would guarantee freeing all freeable
objects that existed at the moment of starting the operation, while not
looping forever for newly allocated objects, but that's not feasible to
track. In the less ideal solution based on thresholds, arguably the
termination guarantee is more important than the exhaustiveness guarantee.
If there are reports of large regression wrt being exhaustive, we can
tune how fast the threshold grows.
[1] https://lore.kernel.org/lkml/20200909152047.27905-1-zangchunxin@bytedance.com/T/#u
[vbabka@suse.cz: avoid undefined shift behaviour]
Link: https://lkml.kernel.org/r/2f034e6f-a753-550a-f374-e4e23899d3d5@suse.cz
Link: https://lkml.kernel.org/r/20210818152239.25502-1-vbabka@suse.cz
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reported-by: Chunxin Zang <zangchunxin@bytedance.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Chris Down <chris@chrisdown.name>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Kefeng Wang <wangkefeng.wang@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Some method is obviously needed to enable reclaim-based migration.
Just like traditional autonuma, there will be some workloads that will
benefit like workloads with more "static" configurations where hot pages
stay hot and cold pages stay cold. If pages come and go from the hot and
cold sets, the benefits of this approach will be more limited.
The benefits are truly workload-based and *not* hardware-based. We do not
believe that there is a viable threshold where certain hardware
configurations should have this mechanism enabled while others do not.
To be conservative, earlier work defaulted to disable reclaim- based
migration and did not include a mechanism to enable it. This proposes add
a new sysfs file
/sys/kernel/mm/numa/demotion_enabled
as a method to enable it.
We are open to any alternative that allows end users to enable this
mechanism or disable it if workload harm is detected (just like
traditional autonuma).
Once this is enabled page demotion may move data to a NUMA node that does
not fall into the cpuset of the allocating process. This could be
construed to violate the guarantees of cpusets. However, since this is an
opt-in mechanism, the assumption is that anyone enabling it is content to
relax the guarantees.
Link: https://lkml.kernel.org/r/20210721063926.3024591-9-ying.huang@intel.com
Link: https://lkml.kernel.org/r/20210715055145.195411-10-ying.huang@intel.com
Signed-off-by: Huang Ying <ying.huang@intel.com>
Originally-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Wei Xu <weixugc@google.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Keith Busch <kbusch@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Yang Shi <shy828301@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
At the moment memcg stats are read in four contexts:
1. memcg stat user interfaces
2. dirty throttling
3. page fault
4. memory reclaim
Currently the kernel flushes the stats for first two cases. Flushing the
stats for remaining two casese may have performance impact. Always
flushing the memcg stats on the page fault code path may negatively
impacts the performance of the applications. In addition flushing in the
memory reclaim code path, though treated as slowpath, can become the
source of contention for the global lock taken for stat flushing because
when system or memcg is under memory pressure, many tasks may enter the
reclaim path.
This patch uses following mechanisms to solve these challenges:
1. Periodically flush the stats from root memcg every 2 seconds. This
will time limit the out of sync stats.
2. Asynchronously flush the stats after fixed number of stat updates.
In the worst case the stat can be out of sync by O(nr_cpus * BATCH) for
2 seconds.
3. For avoiding thundering herd to flush the stats particularly from
the memory reclaim context, introduce memcg local spinlock and let only
one flusher active at a time. This could have been done through
cgroup_rstat_lock lock but that lock is used by other subsystem and for
userspace reading memcg stats. So, it is better to keep flushers
introduced by this patch decoupled from cgroup_rstat_lock. However we
would have to use irqsafe version of rstat flush but that is fine as
this code path will be flushing for whole tree and do the work for
everyone. No one will be waiting for that worker.
[shakeelb@google.com: fix sleep-in-wrong context bug]
Link: https://lkml.kernel.org/r/20210716212137.1391164-2-shakeelb@google.com
Link: https://lkml.kernel.org/r/20210714013948.270662-2-shakeelb@google.com
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Hillf Danton <hdanton@sina.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A recent lockdep report included these lines:
[ 96.177910] 3 locks held by containerd/770:
[ 96.177934] #0: ffff88810815ea28 (&mm->mmap_lock#2){++++}-{3:3},
at: do_user_addr_fault+0x115/0x770
[ 96.177999] #1: ffffffff82915020 (rcu_read_lock){....}-{1:2}, at:
get_swap_device+0x33/0x140
[ 96.178057] #2: ffffffff82955ba0 (fs_reclaim){+.+.}-{0:0}, at:
__fs_reclaim_acquire+0x5/0x30
While it was not useful to that bug report to know where the reclaim lock
had been acquired, it might be useful under other circumstances. Allow
the caller of __fs_reclaim_acquire to specify the instruction pointer to
use.
Link: https://lkml.kernel.org/r/20210719185709.1755149-1-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Omar Sandoval <osandov@fb.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In a debugging session the other day, Rik noticed that node_reclaim()
was missing memstall annotations. This means we'll miss pressure and
lost productivity resulting from reclaim on an overloaded local NUMA
node when vm.zone_reclaim_mode is enabled.
There haven't been any reports, but that's likely because
vm.zone_reclaim_mode hasn't been a commonly used feature recently, and
the intersection between such setups and psi users is probably nil.
But secondary memory such as CXL-connected DIMMS, persistent memory etc,
and the page demotion patches that handle them
(https://lore.kernel.org/lkml/20210401183216.443C4443@viggo.jf.intel.com/)
could soon make this a more common codepath again.
Link: https://lkml.kernel.org/r/20210818152457.35846-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: Rik van Riel <riel@surriel.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
We've noticed occasional OOM killing when memory.low settings are in
effect for cgroups. This is unexpected and undesirable as memory.low is
supposed to express non-OOMing memory priorities between cgroups.
The reason for this is proportional memory.low reclaim. When cgroups
are below their memory.low threshold, reclaim passes them over in the
first round, and then retries if it couldn't find pages anywhere else.
But when cgroups are slightly above their memory.low setting, page scan
force is scaled down and diminished in proportion to the overage, to the
point where it can cause reclaim to fail as well - only in that case we
currently don't retry, and instead trigger OOM.
To fix this, hook proportional reclaim into the same retry logic we have
in place for when cgroups are skipped entirely. This way if reclaim
fails and some cgroups were scanned with diminished pressure, we'll try
another full-force cycle before giving up and OOMing.
[akpm@linux-foundation.org: coding-style fixes]
Link: https://lkml.kernel.org/r/20210817180506.220056-1-hannes@cmpxchg.org
Fixes: 9783aa9917 ("mm, memcg: proportional memory.{low,min} reclaim")
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: Leon Yang <lnyng@fb.com>
Reviewed-by: Rik van Riel <riel@surriel.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Roman Gushchin <guro@fb.com>
Acked-by: Chris Down <chris@chrisdown.name>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: <stable@vger.kernel.org> [5.4+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Patch series "Clean W=1 build warnings for mm/".
This is a janitorial only. During development of a tool to catch build
warnings early to avoid tripping the Intel lkp-robot, I noticed that mm/
is not clean for W=1. This is generally harmless but there is no harm in
cleaning it up. It disrupts git blame a little but on relatively obvious
lines that are unlikely to be git blame targets.
This patch (of 13):
make W=1 generates the following warning for vmscan.c
mm/vmscan.c:1814: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
It is not a kerneldoc comment and isolate_lru_pages() is a static
function. While the detailed comment is nice, it does not need to be
exposed via kernel-doc.
Link: https://lkml.kernel.org/r/20210520084809.8576-1-mgorman@techsingularity.net
Link: https://lkml.kernel.org/r/20210520084809.8576-2-mgorman@techsingularity.net
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Theoretically without the protect from memalloc_noreclaim_save() and
memalloc_noreclaim_restore(), reclaim_pages() can go into the block
I/O layer recursively and deadlock.
Querying 'reclaim_pages' in our kernel crash databases didn't yield
any results. So the deadlock seems unlikely to happen. A possible
explanation is that the only user of reclaim_pages(), i.e.,
MADV_PAGEOUT, is usually called before memory pressure builds up,
e.g., on Android and Chrome OS. Under such a condition, allocations in
the block I/O layer can be fulfilled without diverting to direct
reclaim and therefore the recursion is avoided.
Link: https://lkml.kernel.org/r/20210622074642.785473-1-yuzhao@google.com
Link: https://lkml.kernel.org/r/20210614194727.2684053-1-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Cc: Minchan Kim <minchan@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently the number of deferred objects are per shrinker, but some
slabs, for example, vfs inode/dentry cache are per memcg, this would
result in poor isolation among memcgs.
The deferred objects typically are generated by __GFP_NOFS allocations,
one memcg with excessive __GFP_NOFS allocations may blow up deferred
objects, then other innocent memcgs may suffer from over shrink,
excessive reclaim latency, etc.
For example, two workloads run in memcgA and memcgB respectively,
workload in B is vfs heavy workload. Workload in A generates excessive
deferred objects, then B's vfs cache might be hit heavily (drop half of
caches) by B's limit reclaim or global reclaim.
We observed this hit in our production environment which was running vfs
heavy workload shown as the below tracing log:
<...>-409454 [016] .... 28286961.747146: mm_shrink_slab_start: super_cache_scan+0x0/0x1a0 ffff9a83046f3458:
nid: 1 objects to shrink 3641681686040 gfp_flags GFP_HIGHUSER_MOVABLE|__GFP_ZERO pgs_scanned 1 lru_pgs 15721
cache items 246404277 delta 31345 total_scan 123202138
<...>-409454 [022] .... 28287105.928018: mm_shrink_slab_end: super_cache_scan+0x0/0x1a0 ffff9a83046f3458:
nid: 1 unused scan count 3641681686040 new scan count 3641798379189 total_scan 602
last shrinker return val 123186855
The vfs cache and page cache ratio was 10:1 on this machine, and half of
caches were dropped. This also resulted in significant amount of page
caches were dropped due to inodes eviction.
Make nr_deferred per memcg for memcg aware shrinkers would solve the
unfairness and bring better isolation.
The following patch will add nr_deferred to parent memcg when memcg
offline. To preserve nr_deferred when reparenting memcgs to root, root
memcg needs shrinker_info allocated too.
When memcg is not enabled (!CONFIG_MEMCG or memcg disabled), the
shrinker's nr_deferred would be used. And non memcg aware shrinkers use
shrinker's nr_deferred all the time.
Link: https://lkml.kernel.org/r/20210311190845.9708-10-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Acked-by: Roman Gushchin <guro@fb.com>
Acked-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
