For each accessed chunk a PTE is created. More than 1GiB of PTEs is used
in this way. Remove each PTE after validating a chunk to reduce peak
memory usage.
It is important to only unmap memory that previously mmap()ed, as
unmapping other mappings like the stack, heap or executable mappings will
crash the process.
The mappings read from /proc/self/maps and the return values from mmap()
don't allow a simple correlation due to merging and no guaranteed order.
To correlate the pointers and mappings use prctl(PR_SET_VMA_ANON_NAME).
While it introduces a test dependency, other alternatives would introduce
runtime or development overhead.
Link: https://lkml.kernel.org/r/20250114-virtual_address_range-tests-v4-2-6fd7269934a5@linutronix.de
Fixes: 0104096498 ("selftests/mm: confirm VA exhaustion without reliance on correctness of mmap()")
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Cc: Dev Jain <dev.jain@arm.com>
Cc: kernel test robot <oliver.sang@intel.com>
Cc: Shuah Khan (Samsung OSG) <shuah@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add RWF_DONTCACHE as a read operation flag, which means that any data read
wil be removed from the page cache upon completion. Uses the page cache
to synchronize, and simply prunes folios that were instantiated when the
operation completes. While it would be possible to use private pages for
this, using the page cache as synchronization is handy for a variety of
reasons:
1) No special truncate magic is needed
2) Async buffered reads need some place to serialize, using the page
cache is a lot easier than writing extra code for this
3) The pruning cost is pretty reasonable
and the code to support this is much simpler as a result.
You can think of uncached buffered IO as being the much more attractive
cousin of O_DIRECT - it has none of the restrictions of O_DIRECT. Yes, it
will copy the data, but unlike regular buffered IO, it doesn't run into
the unpredictability of the page cache in terms of reclaim. As an
example, on a test box with 32 drives, reading them with buffered IO looks
as follows:
Reading bs 65536, uncached 0
1s: 145945MB/sec
2s: 158067MB/sec
3s: 157007MB/sec
4s: 148622MB/sec
5s: 118824MB/sec
6s: 70494MB/sec
7s: 41754MB/sec
8s: 90811MB/sec
9s: 92204MB/sec
10s: 95178MB/sec
11s: 95488MB/sec
12s: 95552MB/sec
13s: 96275MB/sec
where it's quite easy to see where the page cache filled up, and
performance went from good to erratic, and finally settles at a much
lower rate. Looking at top while this is ongoing, we see:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7535 root 20 0 267004 0 0 S 3199 0.0 8:40.65 uncached
3326 root 20 0 0 0 0 R 100.0 0.0 0:16.40 kswapd4
3327 root 20 0 0 0 0 R 100.0 0.0 0:17.22 kswapd5
3328 root 20 0 0 0 0 R 100.0 0.0 0:13.29 kswapd6
3332 root 20 0 0 0 0 R 100.0 0.0 0:11.11 kswapd10
3339 root 20 0 0 0 0 R 100.0 0.0 0:16.25 kswapd17
3348 root 20 0 0 0 0 R 100.0 0.0 0:16.40 kswapd26
3343 root 20 0 0 0 0 R 100.0 0.0 0:16.30 kswapd21
3344 root 20 0 0 0 0 R 100.0 0.0 0:11.92 kswapd22
3349 root 20 0 0 0 0 R 100.0 0.0 0:16.28 kswapd27
3352 root 20 0 0 0 0 R 99.7 0.0 0:11.89 kswapd30
3353 root 20 0 0 0 0 R 96.7 0.0 0:16.04 kswapd31
3329 root 20 0 0 0 0 R 96.4 0.0 0:11.41 kswapd7
3345 root 20 0 0 0 0 R 96.4 0.0 0:13.40 kswapd23
3330 root 20 0 0 0 0 S 91.1 0.0 0:08.28 kswapd8
3350 root 20 0 0 0 0 S 86.8 0.0 0:11.13 kswapd28
3325 root 20 0 0 0 0 S 76.3 0.0 0:07.43 kswapd3
3341 root 20 0 0 0 0 S 74.7 0.0 0:08.85 kswapd19
3334 root 20 0 0 0 0 S 71.7 0.0 0:10.04 kswapd12
3351 root 20 0 0 0 0 R 60.5 0.0 0:09.59 kswapd29
3323 root 20 0 0 0 0 R 57.6 0.0 0:11.50 kswapd1
[...]
which is just showing a partial list of the 32 kswapd threads that are
running mostly full tilt, burning ~28 full CPU cores.
If the same test case is run with RWF_DONTCACHE set for the buffered read,
the output looks as follows:
Reading bs 65536, uncached 0
1s: 153144MB/sec
2s: 156760MB/sec
3s: 158110MB/sec
4s: 158009MB/sec
5s: 158043MB/sec
6s: 157638MB/sec
7s: 157999MB/sec
8s: 158024MB/sec
9s: 157764MB/sec
10s: 157477MB/sec
11s: 157417MB/sec
12s: 157455MB/sec
13s: 157233MB/sec
14s: 156692MB/sec
which is just chugging along at ~155GB/sec of read performance. Looking
at top, we see:
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7961 root 20 0 267004 0 0 S 3180 0.0 5:37.95 uncached
8024 axboe 20 0 14292 4096 0 R 1.0 0.0 0:00.13 top
where just the test app is using CPU, no reclaim is taking place outside
of the main thread. Not only is performance 65% better, it's also using
half the CPU to do it.
Link: https://lkml.kernel.org/r/20241220154831.1086649-9-axboe@kernel.dk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Cc: Brian Foster <bfoster@redhat.com>
Cc: Chris Mason <clm@meta.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add a folio_unmap_invalidate() helper, which unmaps and invalidates a
given folio. The caller must already have locked the folio. Embed the
old invalidate_complete_folio2() helper in there as well, as nobody else
calls it.
Use this new helper in invalidate_inode_pages2_range(), rather than
duplicate the code there.
In preparation for using this elsewhere as well, have it take a gfp_t mask
rather than assume GFP_KERNEL is the right choice. This bubbles back to
invalidate_complete_folio2() as well.
Link: https://lkml.kernel.org/r/20241220154831.1086649-7-axboe@kernel.dk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Cc: Brian Foster <bfoster@redhat.com>
Cc: Chris Mason <clm@meta.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "Uncached buffered IO", v8.
5 years ago I posted patches adding support for RWF_UNCACHED, as a way to
do buffered IO that isn't page cache persistent. The approach back then
was to have private pages for IO, and then get rid of them once IO was
done. But that then runs into all the issues that O_DIRECT has, in terms
of synchronizing with the page cache.
So here's a new approach to the same concent, but using the page cache as
synchronization. Due to excessive bike shedding on the naming, this is
now named RWF_DONTCACHE, and is less special in that it's just page cache
IO, except it prunes the ranges once IO is completed.
Why do this, you may ask? The tldr is that device speeds are only getting
faster, while reclaim is not. Doing normal buffered IO can be very
unpredictable, and suck up a lot of resources on the reclaim side. This
leads people to use O_DIRECT as a work-around, which has its own set of
restrictions in terms of size, offset, and length of IO. It's also
inherently synchronous, and now you need async IO as well. While the
latter isn't necessarily a big problem as we have good options available
there, it also should not be a requirement when all you want to do is read
or write some data without caching.
Even on desktop type systems, a normal NVMe device can fill the entire
page cache in seconds. On the big system I used for testing, there's a
lot more RAM, but also a lot more devices. As can be seen in some of the
results in the following patches, you can still fill RAM in seconds even
when there's 1TB of it. Hence this problem isn't solely a "big
hyperscaler system" issue, it's common across the board.
Common for both reads and writes with RWF_DONTCACHE is that they use the
page cache for IO. Reads work just like a normal buffered read would,
with the only exception being that the touched ranges will get pruned
after data has been copied. For writes, the ranges will get writeback
kicked off before the syscall returns, and then writeback completion will
prune the range. Hence writes aren't synchronous, and it's easy to
pipeline writes using RWF_DONTCACHE. Folios that aren't instantiated by
RWF_DONTCACHE IO are left untouched. This means you that uncached IO will
take advantage of the page cache for uptodate data, but not leave anything
it instantiated/created in cache.
File systems need to support this. This patchset adds support for the
generic read path, which covers file systems like ext4. Patches exist to
add support for iomap/XFS and btrfs as well, which sit on top of this
series. If RWF_DONTCACHE IO is attempted on a file system that doesn't
support it, -EOPNOTSUPP is returned. Hence the user can rely on it either
working as designed, or flagging and error if that's not the case. The
intent here is to give the application a sensible fallback path - eg, it
may fall back to O_DIRECT if appropriate, or just live with the fact that
uncached IO isn't available and do normal buffered IO.
Adding "support" to other file systems should be trivial, most of the time
just a one-liner adding FOP_DONTCACHE to the fop_flags in the
file_operations struct, if the file system is using either iomap or the
generic filemap helpers for reading and writing.
Performance results are in patch 8 for reads, and you can find the write
side results in the XFS patch adding support for DONTCACHE writes for XFS:
https://git.kernel.dk/cgit/linux/commit/?h=buffered-uncached-fs.10&id=257e92de795fdff7d7e256501e024fac6da6a7f4
with the tldr being that I see about a 65% improvement in performance for
both, with fully predictable IO times. CPU reduction is substantial as
well, with no kswapd activity at all for reclaim when using uncached IO.
Using it from applications is trivial - just set RWF_DONTCACHE for the
read or write, using pwritev2(2) or preadv2(2). For io_uring, same thing,
just set RWF_DONTCACHE in sqe->rw_flags for a buffered read/write
operation. And that's it.
Patches 1..7 are just prep patches, and should have no functional changes
at all. Patch 8 adds support for the filemap path for RWF_DONTCACHE
reads, and patches 9..12 are just prep patches for supporting the write
side of uncached writes. In the below mentioned branch, there are then
patches to adopt uncached reads and writes for xfs, btrfs, and ext4. The
latter currently relies on bit of a hack for passing whether this is an
uncached write or not through ->write_begin(), which can hopefully go away
once ext4 adopts iomap for buffered writes. I say this is a hack as it's
not the prettiest way to do it, however it is fully solid and will work
just fine.
Passes full xfstests and fsx overnight runs, no issues observed. That
includes the vm running the testing also using RWF_DONTCACHE on the host.
I'll post fsstress and fsx patches for RWF_DONTCACHE separately. As far
as I'm concerned, no further work needs doing here.
And git tree for the patches is here:
https://git.kernel.dk/cgit/linux/log/?h=buffered-uncached.10
with the file system patches on top adding support for xfs/btrfs/ext4
here:
https://git.kernel.dk/cgit/linux/log/?h=buffered-uncached-fs.10
This patch (of 12):
Rather than pass in both the file and position directly from the kiocb,
just take a struct kiocb instead. With the kiocb being passed in, skip
passing in the address_space separately as well. While doing so, move the
ki_flags checking into filemap_create_folio() as well. In preparation for
actually needing the kiocb in the function.
No functional changes in this patch.
Link: https://lkml.kernel.org/r/20241220154831.1086649-1-axboe@kernel.dk
Link: https://lkml.kernel.org/r/20241220154831.1086649-2-axboe@kernel.dk
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Brian Foster <bfoster@redhat.com>
Cc: Chris Mason <clm@meta.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
We replaced a simple put_page() by a putback_active_hugepage() call in
commit 3aaa76e125 ("mm: migrate: hugetlb: putback destination hugepage
to active list"), to set the "active" flag on the dst hugetlb folio.
Nowadays, we decoupled the "active" list from the flag, by calling the
flag "migratable".
Calling "putback" on something that wasn't allocated is weird and not
future proof, especially if we might reach that path when migration failed
and we just want to free the freshly allocated hugetlb folio.
Let's simply handle the migratable flag and the active list flag in
move_hugetlb_state(), where we know that allocation succeeded and already
handle the temporary flag; use a simple folio_put() to return our
reference.
Link: https://lkml.kernel.org/r/20250113131611.2554758-4-david@redhat.com
Signed-off-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Sidhartha Kumar <sidhartha.kumar@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Use the correct kernel-doc character following function parameters or
struct members (':' instead of '-') to eliminate kernel-doc warnings.
kasan.h:509: warning: Function parameter or struct member 'addr' not described in 'kasan_poison'
kasan.h:509: warning: Function parameter or struct member 'size' not described in 'kasan_poison'
kasan.h:509: warning: Function parameter or struct member 'value' not described in 'kasan_poison'
kasan.h:509: warning: Function parameter or struct member 'init' not described in 'kasan_poison'
kasan.h:522: warning: Function parameter or struct member 'addr' not described in 'kasan_unpoison'
kasan.h:522: warning: Function parameter or struct member 'size' not described in 'kasan_unpoison'
kasan.h:522: warning: Function parameter or struct member 'init' not described in 'kasan_unpoison'
kasan.h:539: warning: Function parameter or struct member 'address' not described in 'kasan_poison_last_granule'
kasan.h:539: warning: Function parameter or struct member 'size' not described in 'kasan_poison_last_granule'
Link: https://lkml.kernel.org/r/20250111063249.910975-1-rdunlap@infradead.org
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
In /proc/<pid>/ksm_stat, add two extra ksm involvement items including
KSM_mergeable and KSM_merge_any. It helps administrators to better know
the system's KSM behavior at process level.
ksm_merge_any: yes/no
whether the process'mm is added by prctl() into the candidate list
of KSM or not, and fully enabled at process level.
ksm_mergeable: yes/no
whether any VMAs of the process'mm are currently applicable to KSM.
Purpose
=======
These two items are just to improve the observability of KSM at process
level, so that users can know if a certain process has enabled KSM.
For example, if without these two items, when we look at
/proc/<pid>/ksm_stat and there's no merging pages found, We are not sure
whether it is because KSM was not enabled or because KSM did not
successfully merge any pages.
Although "mg" in /proc/<pid>/smaps indicate VM_MERGEABLE, it's opaque
and not very obvious for non professionals.
[akpm@linux-foundation.org: wording tweaks, per David and akpm]
Link: https://lkml.kernel.org/r/20250110174034304QOb8eDoqtFkp3_t8mqnqc@zte.com.cn
Signed-off-by: xu xin <xu.xin16@zte.com.cn>
Acked-by: David Hildenbrand <david@redhat.com>
Tested-by: Mario Casquero <mcasquer@redhat.com>
Cc: Wang Yaxin <wang.yaxin@zte.com.cn>
Cc: Yang Yang <yang.yang29@zte.com.cn>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The existing logic uses strnlen_user() to calculate the length of the
memfd name from userspace and then copies the string into a buffer using
copy_from_user(). This is error-prone, as the string length could have
changed between the time when it was calculated and when the string was
copied. The existing logic handles this by ensuring that the last byte in
the buffer is the terminating zero.
This handling is contrived and can better be handled by using
strncpy_from_user(), which gets the length of the string and copies it in
one shot. Therefore, simplify the logic for copying the memfd name by
using strncpy_from_user().
No functional change.
Link: https://lkml.kernel.org/r/20250110165904.3437374-3-isaacmanjarres@google.com
Signed-off-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: John Stultz <jstultz@google.com>
Cc: Kalesh Singh <kaleshsingh@google.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "Cleanup for memfd_create()", v4.
memfd_create() handles all of its logic in a single function. Some of the
logic in the function is also somewhat contrived (i.e. copying the memfd
name from userpace).
This series aims to cleanup memfd_create() by splitting out the logic into
helper functions, and simplifying the memfd name copying to make the code
easier to follow.
This has no intended functional changes.
Thank you Alice and Lorenzo for reviewing v3 of this series and for your
feedback!
This patch (of 2):
memfd_create() is a pretty busy function that could be easier to read if
some of the logic was split out into helper functions.
Therefore, split the flags sanitization, name allocation, and file
structure allocation into their own helper functions.
No functional change.
Link: https://lkml.kernel.org/r/20250110165904.3437374-1-isaacmanjarres@google.com
Link: https://lkml.kernel.org/r/20250110165904.3437374-2-isaacmanjarres@google.com
Signed-off-by: Isaac J. Manjarres <isaacmanjarres@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: Isaac J. Manjarres <isaacmanjarres@google.com>
Cc: John Stultz <jstultz@google.com>
Cc: Kalesh Singh <kaleshsingh@google.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Add a DAMON monitoring intervals tuning example that contains output from
a demonstration of the guide on a real server workload system. The
example with real world numbers will help users better understanding the
guide instructions and what outputs they can expect and verify. Those
will again help finding the rooms for improvements on the guide.
Link: https://lkml.kernel.org/r/20250110185232.54907-3-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: Honggyu Kim <honggyu.kim@sk.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Yunjeong Mun <yunjeong.mun@sk.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Count the accessed bits from PTEs mapping the same large folio as one
access rather than multiple accesses.
The last patch changed how folios accessed through page tables are
promoted: rather than getting promoted after the accessed bit is cleared
for the first time, a folio only gets promoted thereafter. Counting the
accessed bits from the same large folio as multiple accesses can cause
that folio to be promoted prematurely, which in turn can cause
overprotection of single-use large folios.
This patch reduced the sys time of the kernel compilation by 95% CI [2,
5]% on Altra M128-30 with 3GB DRAM, 12GB zram, 16KB THPs and -j32.
Link: https://lkml.kernel.org/r/20241231043538.4075764-8-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: Barry Song <v-songbaohua@oppo.com>
Tested-by: Kalesh Singh <kaleshsingh@google.com>
Cc: Bharata B Rao <bharata@amd.com>
Cc: David Stevens <stevensd@chromium.org>
Cc: Kairui Song <kasong@tencent.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
With the aging feedback no longer considering the distribution of folios
in each generation, rework workingset protection to better distribute
folios across MAX_NR_GENS. This is achieved by reusing PG_workingset and
PG_referenced/LRU_REFS_FLAGS in a slightly different way.
For folios accessed multiple times through file descriptors, make
lru_gen_inc_refs() set additional bits of LRU_REFS_WIDTH in folio->flags
after PG_referenced, then PG_workingset after LRU_REFS_WIDTH. After all
its bits are set, i.e., LRU_REFS_FLAGS|BIT(PG_workingset), a folio is
lazily promoted into the second oldest generation in the eviction path.
And when folio_inc_gen() does that, it clears LRU_REFS_FLAGS so that
lru_gen_inc_refs() can start over. For this case, LRU_REFS_MASK is only
valid when PG_referenced is set.
For folios accessed multiple times through page tables, folio_update_gen()
from a page table walk or lru_gen_set_refs() from a rmap walk sets
PG_referenced after the accessed bit is cleared for the first time.
Thereafter, those two paths set PG_workingset and promote folios to the
youngest generation. Like folio_inc_gen(), when folio_update_gen() does
that, it also clears PG_referenced. For this case, LRU_REFS_MASK is not
used.
For both of the cases, after PG_workingset is set on a folio, it remains
until this folio is either reclaimed, or "deactivated" by
lru_gen_clear_refs(). It can be set again if lru_gen_test_recent()
returns true upon a refault.
When adding folios to the LRU lists, lru_gen_folio_seq() distributes
them as follows:
+---------------------------------+---------------------------------+
| Accessed thru page tables | Accessed thru file descriptors |
+---------------------------------+---------------------------------+
| PG_active (set while isolated) | |
+----------------+----------------+----------------+----------------+
| PG_workingset | PG_referenced | PG_workingset | LRU_REFS_FLAGS |
+---------------------------------+---------------------------------+
|<--------- MIN_NR_GENS --------->| |
|<-------------------------- MAX_NR_GENS -------------------------->|
After this patch, some typical client and server workloads showed
improvements under heavy memory pressure. For example, Python TPC-C,
which was used to benchmark a different approach [1] to better detect
refault distances, showed a significant decrease in total refaults:
Before After Change
Time (seconds) 10801 10801 0%
Executed (transactions) 41472 43663 +5%
workingset_nodes 109070 120244 +10%
workingset_refault_anon 5019627 7281831 +45%
workingset_refault_file 1294678786 554855564 -57%
workingset_refault_total 1299698413 562137395 -57%
[1] https://lore.kernel.org/20230920190244.16839-1-ryncsn@gmail.com/
Link: https://lkml.kernel.org/r/20241231043538.4075764-7-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: Kairui Song <kasong@tencent.com>
Closes: https://lore.kernel.org/CAOUHufahuWcKf5f1Sg3emnqX+cODuR=2TQo7T4Gr-QYLujn4RA@mail.gmail.com/
Tested-by: Kalesh Singh <kaleshsingh@google.com>
Cc: Barry Song <v-songbaohua@oppo.com>
Cc: Bharata B Rao <bharata@amd.com>
Cc: David Stevens <stevensd@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
With anon and file min_seq being able to move independently, rework type
selection so that it is based on the total refaults from all tiers of each
type. Also allow a type to be selected until that type reaches
MIN_NR_GENS, regardless of whether that type has a larger min_seq or not,
to accommodate extreme swappiness.
Since some tiers of a selected type can have higher refaults than the
first tier of the other type, use a less larger gain factor 2:3 instead of
1:2, in order for those tiers in the selected type to be better protected.
As an intermediate step to the final optimization, this change by itself
should not have userspace-visiable effects beyond performance.
Link: https://lkml.kernel.org/r/20241231043538.4075764-5-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: David Stevens <stevensd@chromium.org>
Tested-by: Kalesh Singh <kaleshsingh@google.com>
Cc: Barry Song <v-songbaohua@oppo.com>
Cc: Bharata B Rao <bharata@amd.com>
Cc: Kairui Song <kasong@tencent.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The aging feedback is based on both the number of generations and the
distribution of folios in each generation. The number of generations is
currently the distance between max_seq and anon min_seq. This is because
anon min_seq is not allowed to move past file min_seq. The rationale for
that is that file is always evictable whereas anon is not. However, for
use cases where anon is a lot cheaper than file:
1. Anon in the second oldest generation can be a better choice than
file in the oldest generation.
2. A large amount of file in the oldest generation can skew the
distribution, making should_run_aging() return false negative.
Allow anon and file min_seq to move independently, and use solely the
number of generations as the feedback for aging. Specifically, when both
anon and file are evictable, anon min_seq can now be greater than file
min_seq, and therefore the number of generations becomes the distance
between max_seq and min(min_seq[0],min_seq[1]). And should_run_aging()
returns true if and only if the number of generations is less than
MAX_NR_GENS.
As the first step to the final optimization, this change by itself should
not have userspace-visiable effects beyond performance. The next twos
patch will take advantage of this change; the last patch in this series
will better distribute folios across MAX_NR_GENS.
[yuzhao@google.com: restore behaviour for systems with swappiness == 200]
Link: https://lkml.kernel.org/r/Z4S3-aJy5dj9tBTk@google.com
Link: https://lkml.kernel.org/r/20241231043538.4075764-4-yuzhao@google.com
Signed-off-by: Yu Zhao <yuzhao@google.com>
Reported-by: David Stevens <stevensd@chromium.org>
Tested-by: Kalesh Singh <kaleshsingh@google.com>
Cc: Barry Song <v-songbaohua@oppo.com>
Cc: Bharata B Rao <bharata@amd.com>
Cc: Kairui Song <kasong@tencent.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Do not shuffle a folio in the deactivation paths if it is already in the
oldest generation. This reduces the LRU lock contention.
Before this patch, the contention is reproducible by FIO, e.g.,
fio -filename=/dev/nvme1n1p2 -direct=0 -thread -size=1024G \
-rwmixwrite=30 --norandommap --randrepeat=0 -ioengine=sync \
-bs=4k -numjobs=400 -runtime=25000 --time_based \
-group_reporting -name=mglru
98.96%--_raw_spin_lock_irqsave
folio_lruvec_lock_irqsave
|
--98.78%--folio_batch_move_lru
|
--98.63%--deactivate_file_folio
mapping_try_invalidate
invalidate_mapping_pages
invalidate_bdev
blkdev_common_ioctl
blkdev_ioctl
After this patch, deactivate_file_folio() bails out early without taking
the LRU lock.
A side effect is that a folio can be left at the head of the oldest
generation, rather than the tail. If reclaim happens at the same time, it
cannot reclaim this folio immediately. Since there is no known
correlation between truncation and reclaim, this side effect is considered
insignificant.
Link: https://lkml.kernel.org/r/20241231043538.4075764-3-yuzhao@google.com
Reported-by: Bharata B Rao <bharata@amd.com>
Closes: https://lore.kernel.org/CAOUHufawNerxqLm7L9Yywp3HJFiYVrYO26ePUb1jH-qxNGWzyA@mail.gmail.com/
Signed-off-by: Yu Zhao <yuzhao@google.com>
Tested-by: Kalesh Singh <kaleshsingh@google.com>
Cc: Barry Song <v-songbaohua@oppo.com>
Cc: David Stevens <stevensd@chromium.org>
Cc: Kairui Song <kasong@tencent.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
