Define yet another DAMOS filter type, YOUNG. Like anon and memcg, the
type of filter will be applied to each page in the memory region, and see
if the page is accessed since the last check. Based on the 'matching'
parameter, the page is filtered out or in.
Note that this commit is adding only the type definition. The
implementation should be made by DAMON operations sets. A commit for the
implementation on 'paddr' DAMON operations set will follow.
Link: https://lkml.kernel.org/r/20240426195247.100306-4-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Tested-by: Honggyu Kim <honggyu.kim@sk.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
damon_pa_mkold() receives physical address, get the folio covering the
address, and makes the folio as old. A following commit will reuse the
internal logic for marking a given folio as old. To avoid duplication of
the code, split the internal logic. Also, change the rmap walker
function's name from __damon_pa_mkold() to damon_folio_mkold_one(),
following the change of the caller's name and the naming rule that more
commonly used by other rmap walkers.
Link: https://lkml.kernel.org/r/20240426195247.100306-3-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Tested-by: Honggyu Kim <honggyu.kim@sk.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm/damon: add a DAMOS filter type for page granularity
access recheck".
DAMON provides its best-effort accuracy-overhead tradeoff under the
user-defined ranges of acceptable level of the monitoring accuracy and
overhead. A recent discussion for tiered memory management support from
DAMON[1] concluded that finding memory regions of specific access pattern
with low overhead despite of low accuracy via DAMON first, and then double
checking the access of the region again in a finer (e.g., page)
granularity could be a useful strategy for some DAMOS schemes.
Add a new type of DAMOS filter, namely 'young' for such a case. It checks
each page of DAMOS target region is accessed since the last check, and
filters it out or in if 'matching' parameter is 'true' or 'false',
respectively.
Because this is a filter type that applied in page granularity, the
support depends on DAMON operations set, similar to 'anon' and 'memcg'
DAMOS filter types. Implement the support on the DAMON operations set for
the physical address space, 'paddr', since one of the expected usages[1]
is based on the physical address space.
[1] https://lore.kernel.org/r/20240227235121.153277-1-sj@kernel.org
This patch (of 7):
damon_pa_young() receives physical address, get the folio covering the
address, and show if the folio is accessed since the last check. A
following commit will reuse the internal logic for checking access to a
given folio. To avoid duplication of the code, split the internal logic.
Also, change the rmap walker function's name from __damon_pa_young() to
damon_folio_young_one(), following the change of the caller's name and the
naming rule that more commonly used by other rmap walkers.
Link: https://lkml.kernel.org/r/20240426195247.100306-1-sj@kernel.org
Link: https://lkml.kernel.org/r/20240426195247.100306-2-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Tested-by: Honggyu Kim <honggyu.kim@sk.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Instead of checking the anon_vma early in the fault path where all page
faults pay the cost, delay it until we know we're going to need the
anon_vma to be filled in. This will have a slight negative effect on the
first fault in an anonymous VMA, but it shortens every other page fault.
It also makes the code slightly cleaner as the anon and file backed fault
handling look more similar.
The Intel kernel test bot reports a 3x improvement in vm-scalability
throughput with the small-allocs-mt test. This is clearly an extreme
situation that won't be replicated in any real-world workload, but it's a
nice win.
https://lore.kernel.org/all/202404261055.c5e24608-oliver.sang@intel.com/
Link: https://lkml.kernel.org/r/20240426144506.1290619-3-willy@infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Suren Baghdasaryan <surenb@google.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Jann Horn <jannh@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
wb_calc_thresh() is calculating wb's share of bg_thresh in the global
domain. However in case of cgroup writeback this is not the right
thing to do. Consider the following domain hierarchy:
global domain (> 20G)
/ \
cgroup1 (10G) cgroup2 (10G)
| |
bdi wb1 wb2
and assume wb1 and wb2 have the same bandwidth and the background
threshold is set at 10%. The bg_thresh of cgroup1 and cgroup2 is going
to be 1G. Now because wb_calc_thresh(mdtc->wb, mdtc->bg_thresh)
calculates per-wb threshold in the global domain as (wb bandwidth) /
(domain bandwidth) it returns bg_thresh for wb1 as 0.5G although it has
nobody to compete against in cgroup1.
Fix the problem by calculating wb's share of bg_thresh in the cgroup
domain.
Test as following:
/* make it easier to observe the issue */
echo 300000 > /proc/sys/vm/dirty_expire_centisecs
echo 100 > /proc/sys/vm/dirty_writeback_centisecs
/* run fio in wb1 */
cd /sys/fs/cgroup
echo "+memory +io" > cgroup.subtree_control
mkdir group1
cd group1
echo 10G > memory.high
echo 10G > memory.max
echo $$ > cgroup.procs
mkfs.ext4 -F /dev/vdb
mount /dev/vdb /bdi1/
fio -name test -filename=/bdi1/file -size=600M -ioengine=libaio -bs=4K \
-iodepth=1 -rw=write -direct=0 --time_based -runtime=600 -invalidate=0
/* run fio in wb2 with a new shell */
cd /sys/fs/cgroup
mkdir group2
cd group2
echo 10G > memory.high
echo 10G > memory.max
echo $$ > cgroup.procs
mkfs.ext4 -F /dev/vdc
mount /dev/vdc /bdi2/
fio -name test -filename=/bdi2/file -size=600M -ioengine=libaio -bs=4K \
-iodepth=1 -rw=write -direct=0 --time_based -runtime=600 -invalidate=0
Before fix, the wrttien pages of wb1 and wb2 reported from
toos/writeback/wb_monitor.py keep growing. After fix, rare written pages
are accumulated.
There is no obvious change in fio result.
[jack@suse.cz: changelog rewording]
Link: https://lkml.kernel.org/r/20240425131724.36778-3-shikemeng@huaweicloud.com
Fixes: 74d3694433 ("writeback: Fix performance regression in wb_over_bg_thresh()")
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Howard Cochran <hcochran@kernelspring.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Miklos Szeredi <mszeredi@redhat.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add wb_monitor.py script to monitor writeback information on backing dev
which makes it easier and more convenient to observe writeback behaviors
of running system.
The wb_monitor.py script is written based on wq_monitor.py.
Following domain hierarchy is tested:
global domain (320G)
/ \
cgroup domain1(10G) cgroup domain2(10G)
| |
bdi wb1 wb2
The wb_monitor.py script output is as following:
./wb_monitor.py 252:16 -c
writeback reclaimable dirtied written avg_bw
252:16_1 0 0 0 0 102400
252:16_4284 672 820064 9230368 8410304 685612
252:16_4325 896 819840 10491264 9671648 652348
252:16 1568 1639904 19721632 18081952 1440360
writeback reclaimable dirtied written avg_bw
252:16_1 0 0 0 0 102400
252:16_4284 672 820064 9230368 8410304 685612
252:16_4325 896 819840 10491264 9671648 652348
252:16 1568 1639904 19721632 18081952 1440360
...
Link: https://lkml.kernel.org/r/20240423034643.141219-5-shikemeng@huaweicloud.com
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Suggested-by: Tejun Heo <tj@kernel.org>
Cc: Brian Foster <bfoster@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: SeongJae Park <sj@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "Improve visibility of writeback", v5.
This series tries to improve visilibity of writeback. Patch 1 make
/sys/kernel/debug/bdi/xxx/stats show writeback info of whole bdi instead
of only writeback info in root cgroup. Patch 2 add a new debug file
/sys/kernel/debug/bdi/xxx/wb_stats to show per wb writeback info. Patch 3
add wb_monitor.py to monitor basic writeback info of running system, more
info could be added on demand. Patch 4 is a random cleanup. More details
can be found in respective patches.
Following domain hierarchy is tested:
global domain (320G)
/ \
cgroup domain1(10G) cgroup domain2(10G)
| |
bdi wb1 wb2
/* all writeback info of bdi is successfully collected */
cat stats
BdiWriteback: 4704 kB
BdiReclaimable: 1294496 kB
BdiDirtyThresh: 204208088 kB
DirtyThresh: 195259944 kB
BackgroundThresh: 32503588 kB
BdiDirtied: 48519296 kB
BdiWritten: 47225696 kB
BdiWriteBandwidth: 1173892 kBps
b_dirty: 1
b_io: 0
b_more_io: 1
b_dirty_time: 0
bdi_list: 1
state: 1
/* per wb writeback info of bdi is collected */
cat /sys/kernel/debug/bdi/252:16/wb_stats
WbCgIno: 1
WbWriteback: 0 kB
WbReclaimable: 0 kB
WbDirtyThresh: 0 kB
WbDirtied: 0 kB
WbWritten: 0 kB
WbWriteBandwidth: 102400 kBps
b_dirty: 0
b_io: 0
b_more_io: 0
b_dirty_time: 0
state: 1
WbCgIno: 4208
WbWriteback: 59808 kB
WbReclaimable: 676480 kB
WbDirtyThresh: 6004624 kB
WbDirtied: 23348192 kB
WbWritten: 22614592 kB
WbWriteBandwidth: 593204 kBps
b_dirty: 1
b_io: 1
b_more_io: 0
b_dirty_time: 0
state: 7
WbCgIno: 4249
WbWriteback: 144256 kB
WbReclaimable: 432096 kB
WbDirtyThresh: 6004344 kB
WbDirtied: 25727744 kB
WbWritten: 25154752 kB
WbWriteBandwidth: 577904 kBps
b_dirty: 0
b_io: 1
b_more_io: 0
b_dirty_time: 0
state: 7
The wb_monitor.py script output is as following:
./wb_monitor.py 252:16 -c
writeback reclaimable dirtied written avg_bw
252:16_1 0 0 0 0 102400
252:16_4284 672 820064 9230368 8410304 685612
252:16_4325 896 819840 10491264 9671648 652348
252:16 1568 1639904 19721632 18081952 1440360
writeback reclaimable dirtied written avg_bw
252:16_1 0 0 0 0 102400
252:16_4284 672 820064 9230368 8410304 685612
252:16_4325 896 819840 10491264 9671648 652348
252:16 1568 1639904 19721632 18081952 1440360
...
This patch (of 5):
/sys/kernel/debug/bdi/xxx/stats is supposed to show writeback information
of whole bdi, but only writeback information of bdi in root cgroup is
collected. So writeback information in non-root cgroup are missing now.
To be more specific, considering following case:
/* create writeback cgroup */
cd /sys/fs/cgroup
echo "+memory +io" > cgroup.subtree_control
mkdir group1
cd group1
echo $$ > cgroup.procs
/* do writeback in cgroup */
fio -name test -filename=/dev/vdb ...
/* get writeback info of bdi */
cat /sys/kernel/debug/bdi/xxx/stats
The cat result unexpectedly implies that there is no writeback on target
bdi.
Fix this by collecting stats of all wb in bdi instead of only wb in
root cgroup.
Following domain hierarchy is tested:
global domain (320G)
/ \
cgroup domain1(10G) cgroup domain2(10G)
| |
bdi wb1 wb2
/* all writeback info of bdi is successfully collected */
cat stats
BdiWriteback: 2912 kB
BdiReclaimable: 1598464 kB
BdiDirtyThresh: 167479028 kB
DirtyThresh: 195038532 kB
BackgroundThresh: 32466728 kB
BdiDirtied: 19141696 kB
BdiWritten: 17543456 kB
BdiWriteBandwidth: 1136172 kBps
b_dirty: 2
b_io: 0
b_more_io: 1
b_dirty_time: 0
bdi_list: 1
state: 1
Link: https://lkml.kernel.org/r/20240423034643.141219-1-shikemeng@huaweicloud.com
Link: https://lkml.kernel.org/r/20240423034643.141219-2-shikemeng@huaweicloud.com
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: Brian Foster <bfoster@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: David Sterba <dsterba@suse.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: SeongJae Park <sj@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
We want to limit the use of page_mapcount() to places where absolutely
required, to prepare for kernel configs where we won't keep track of
per-page mapcounts in large folios.
khugepaged is one of the remaining "more challenging" page_mapcount()
users, but we might be able to move away from page_mapcount() without
resulting in a significant behavior change that would warrant
special-casing based on kernel configs.
In 2020, we first added support to khugepaged for collapsing COW-shared
pages via commit 9445689f3b ("khugepaged: allow to collapse a page
shared across fork"), followed by support for collapsing PTE-mapped THP in
commit 5503fbf2b0 ("khugepaged: allow to collapse PTE-mapped compound
pages") and limiting the memory waste via the "page_count() > 1" check in
commit 71a2c112a0 ("khugepaged: introduce 'max_ptes_shared' tunable").
As a default, khugepaged will allow up to half of the PTEs to map shared
pages: where page_mapcount() > 1. MADV_COLLAPSE ignores the khugepaged
setting.
khugepaged does currently not care about swapcache page references, and
does not check under folio lock: so in some corner cases the "shared vs.
exclusive" detection might be a bit off, making us detect "exclusive" when
it's actually "shared".
Most of our anonymous folios in the system are usually exclusive. We
frequently see sharing of anonymous folios for a short period of time,
after which our short-lived suprocesses either quit or exec().
There are some famous examples, though, where child processes exist for a
long time, and where memory is COW-shared with a lot of processes
(webservers, webbrowsers, sshd, ...) and COW-sharing is crucial for
reducing the memory footprint. We don't want to suddenly change the
behavior to result in a significant increase in memory waste.
Interestingly, khugepaged will only collapse an anonymous THP if at least
one PTE is writable. After fork(), that means that something (usually a
page fault) populated at least a single exclusive anonymous THP in that
PMD range.
So ... what happens when we switch to "is this folio mapped shared"
instead of "is this page mapped shared" by using
folio_likely_mapped_shared()?
For "not-COW-shared" folios, small folios and for THPs (large folios) that
are completely mapped into at least one process, switching to
folio_likely_mapped_shared() will not result in a change.
We'll only see a change for COW-shared PTE-mapped THPs that are partially
mapped into all involved processes.
There are two cases to consider:
(A) folio_likely_mapped_shared() returns "false" for a PTE-mapped THP
If the folio is detected as exclusive, and it actually is exclusive,
there is no change: page_mapcount() == 1. This is the common case
without fork() or with short-lived child processes.
folio_likely_mapped_shared() might currently still detect a folio as
exclusive although it is shared (false negatives): if the first page is
not mapped multiple times and if the average per-page mapcount is smaller
than 1, implying that (1) the folio is partially mapped and (2) if we are
responsible for many mapcounts by mapping many pages others can't
("mostly exclusive") (3) if we are not responsible for many mapcounts by
mapping little pages ("mostly shared") it won't make a big impact on the
end result.
So while we might now detect a page as "exclusive" although it isn't,
it's not expected to make a big difference in common cases.
(B) folio_likely_mapped_shared() returns "true" for a PTE-mapped THP
folio_likely_mapped_shared() will never detect a large anonymous folio
as shared although it is exclusive: there are no false positives.
If we detect a THP as shared, at least one page of the THP is mapped by
another process. It could well be that some pages are actually exclusive.
For example, our child processes could have unmapped/COW'ed some pages
such that they would now be exclusive to out process, which we now
would treat as still-shared.
Examples:
(1) Parent maps all pages of a THP, child maps some pages. We detect
all pages in the parent as shared although some are actually
exclusive.
(2) Parent maps all but some page of a THP, child maps the remainder.
We detect all pages of the THP that the parent maps as shared
although they are all exclusive.
In (1) we wouldn't collapse a THP right now already: no PTE
is writable, because a write fault would have resulted in COW of a
single page and the parent would no longer map all pages of that THP.
For (2) we would have collapsed a THP in the parent so far, now we
wouldn't as long as the child process is still alive: unless the child
process unmaps the remaining THP pages or we decide to split that THP.
Possibly, the child COW'ed many pages, meaning that it's likely that
we can populate a THP for our child first, and then for our parent.
For (2), we are making really bad use of the THP in the first
place (not even mapped completely in at least one process). If the
THP would be completely partially mapped, it would be on the deferred
split queue where we would split it lazily later.
For short-running child processes, we don't particularly care. For
long-running processes, the expectation is that such scenarios are
rather rare: further, a THP might be best placed if most data in the
PMD range is actually written, implying that we'll have to COW more
pages first before khugepaged would collapse it.
To summarize, in the common case, this change is not expected to matter
much. The more common application of khugepaged operates on exclusive
pages, either before fork() or after a child quit.
Can we improve (A)? Yes, if we implement more precise tracking of "mapped
shared" vs. "mapped exclusively", we could get rid of the false negatives
completely.
Can we improve (B)? We could count how many pages of a large folio we map
inside the current page table and detect that we are responsible for most
of the folio mapcount and conclude "as good as exclusive", which might
help in some cases. ... but likely, some other mechanism should detect
that the THP is not a good use in the scenario (not even mapped completely
in a single process) and try splitting that folio lazily etc.
We'll move the folio_test_anon() check before our "shared" check, so we
might get more expressive results for SCAN_EXCEED_SHARED_PTE: this order
of checks now matches the one in __collapse_huge_page_isolate(). Extend
documentation.
Link: https://lkml.kernel.org/r/20240424122630.495788-1-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Yang Shi <yang.shi@linux.alibaba.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
A data-race issue in memcg rstat occurs when two distinct code paths
access the same 4-byte region concurrently. KCSAN detection triggers the
following BUG as a result.
BUG: KCSAN: data-race in __count_memcg_events / mem_cgroup_css_rstat_flush
write to 0xffffe8ffff98e300 of 4 bytes by task 5274 on cpu 17:
mem_cgroup_css_rstat_flush (mm/memcontrol.c:5850)
cgroup_rstat_flush_locked (kernel/cgroup/rstat.c:243 (discriminator 7))
cgroup_rstat_flush (./include/linux/spinlock.h:401 kernel/cgroup/rstat.c:278)
mem_cgroup_flush_stats.part.0 (mm/memcontrol.c:767)
memory_numa_stat_show (mm/memcontrol.c:6911)
<snip>
read to 0xffffe8ffff98e300 of 4 bytes by task 410848 on cpu 27:
__count_memcg_events (mm/memcontrol.c:725 mm/memcontrol.c:962)
count_memcg_event_mm.part.0 (./include/linux/memcontrol.h:1097 ./include/linux/memcontrol.h:1120)
handle_mm_fault (mm/memory.c:5483 mm/memory.c:5622)
<snip>
value changed: 0x00000029 -> 0x00000000
The race occurs because two code paths access the same "stats_updates"
location. Although "stats_updates" is a per-CPU variable, it is remotely
accessed by another CPU at
cgroup_rstat_flush_locked()->mem_cgroup_css_rstat_flush(), leading to the
data race mentioned.
Considering that memcg_rstat_updated() is in the hot code path, adding a
lock to protect it may not be desirable, especially since this variable
pertains solely to statistics.
Therefore, annotating accesses to stats_updates with READ/WRITE_ONCE() can
prevent KCSAN splats and potential partial reads/writes.
Link: https://lkml.kernel.org/r/20240424125940.2410718-1-leitao@debian.org
Fixes: 9cee7e8ef3 ("mm: memcg: optimize parent iteration in memcg_rstat_updated()")
Signed-off-by: Breno Leitao <leitao@debian.org>
Suggested-by: Shakeel Butt <shakeel.butt@linux.dev>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Shakeel Butt <shakeel.butt@linux.dev>
Reviewed-by: Yosry Ahmed <yosryahmed@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add userfaultfd_wp() check in vmf_orig_pte_uffd_wp() to avoid the
unnecessary FAULT_FLAG_ORIG_PTE_VALID check/pte_marker_entry_uffd_wp() in
most pagefault, note, the function vmf_orig_pte_uffd_wp() is not inlined
in the two kernel versions, the difference is shown below,
perf date,
perf report -i perf.data.before | grep vmf
0.17% 0.13% lat_pagefault [kernel.kallsyms] [k] vmf_orig_pte_uffd_wp.part.0.isra.0
perf report -i perf.data.after | grep vmf
lat_pagefault -W 5 -N 5 /tmp/XXX
latency before after diff
average(8 tests) 0.262675 0.2600375 -0.0026375
Although it's a small, but the uffd_wp is a new feature than previous
kernel, when the vma is not registered with UFFD_WP, let's avoid to
execute the new logical, also adding __always_inline attribute to
vmf_orig_pte_uffd_wp(), which make set_pte_range() only check VM_UFFD_WP
flags without the function call. In addition, directly call the
vmf_orig_pte_uffd_wp() in do_anonymous_page() and set_pte_range() to save
an uffd_wp variable.
Link: https://lkml.kernel.org/r/20240422030039.3293568-1-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
