When CONFIG_DAMON_SYSFS is disabled, the selftests fail with the following
outputs,
not ok 2 selftests: damon: sysfs_update_schemes_tried_regions_wss_estimation.py # exit=1
not ok 3 selftests: damon: damos_quota.py # exit=1
not ok 4 selftests: damon: damos_quota_goal.py # exit=1
not ok 5 selftests: damon: damos_apply_interval.py # exit=1
not ok 6 selftests: damon: damos_tried_regions.py # exit=1
not ok 7 selftests: damon: damon_nr_regions.py # exit=1
not ok 11 selftests: damon: sysfs_update_schemes_tried_regions_hang.py # exit=1
The root cause of this issue is that all the testcases above do not check
the sysfs interface of DAMON whether it exists or not. With this patch
applied, all the testcases above now pass successfully.
Link: https://lkml.kernel.org/r/20250531093937.1555159-1-lienze@kylinos.cn
Signed-off-by: Enze Li <lienze@kylinos.cn>
Reviewed-by: SeongJae Park <sj@kernel.org>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
On systems with NUMA balancing enabled, it has been found that tracking
task activities resulting from NUMA balancing is beneficial. NUMA
balancing employs two mechanisms for task migration: one is to migrate
a task to an idle CPU within its preferred node, and the other is to
swap tasks located on different nodes when they are on each other's
preferred nodes.
The kernel already provides NUMA page migration statistics in
/sys/fs/cgroup/mytest/memory.stat and /proc/{PID}/sched. However, it
lacks statistics regarding task migration and swapping. Therefore,
relevant counts for task migration and swapping should be added.
The following two new fields:
numa_task_migrated
numa_task_swapped
will be shown in /sys/fs/cgroup/{GROUP}/memory.stat, /proc/{PID}/sched
and /proc/vmstat.
Introducing both per-task and per-memory cgroup (memcg) NUMA balancing
statistics facilitates a rapid evaluation of the performance and
resource utilization of the target workload. For instance, users can
first identify the container with high NUMA balancing activity and then
further pinpoint a specific task within that group, and subsequently
adjust the memory policy for that task. In short, although it is
possible to iterate through /proc/$pid/sched to locate the problematic
task, the introduction of aggregated NUMA balancing activity for tasks
within each memcg can assist users in identifying the task more
efficiently through a divide-and-conquer approach.
As Libo Chen pointed out, the memcg event relies on the text names in
vmstat_text, and /proc/vmstat generates corresponding items based on
vmstat_text. Thus, the relevant task migration and swapping events
introduced in vmstat_text also need to be populated by
count_vm_numa_event(), otherwise these values are zero in /proc/vmstat.
In theory, task migration and swap events are part of the scheduler's
activities. The reason for exposing them through the
memory.stat/vmstat interface is that we already have NUMA balancing
statistics in memory.stat/vmstat, and these events are closely related
to each other. Following Shakeel's suggestion, we describe the
end-to-end flow/story of all these events occurring on a timeline for
future reference:
The goal of NUMA balancing is to co-locate a task and its memory pages
on the same NUMA node. There are two strategies: migrate the pages to
the task's node, or migrate the task to the node where its pages
reside.
Suppose a task p1 is running on Node 0, but its pages are located on
Node 1. NUMA page fault statistics for p1 reveal its "page footprint"
across nodes. If NUMA balancing detects that most of p1's pages are on
Node 1:
1.Page Migration Attempt:
The Numa balance first tries to migrate p1's pages to Node 0.
The numa_page_migrate counter increments.
2.Task Migration Strategies:
After the page migration finishes, Numa balance checks every
1 second to see if p1 can be migrated to Node 1.
Case 2.1: Idle CPU Available
If Node 1 has an idle CPU, p1 is directly scheduled there. This
event is logged as numa_task_migrated.
Case 2.2: No Idle CPU (Task Swap)
If all CPUs on Node1 are busy, direct migration could cause CPU
contention or load imbalance. Instead: The Numa balance selects a
candidate task p2 on Node 1 that prefers Node 0 (e.g., due to its own
page footprint). p1 and p2 are swapped. This cross-node swap is
recorded as numa_task_swapped.
Link: https://lkml.kernel.org/r/d00edb12ba0f0de3c5222f61487e65f2ac58f5b1.1748493462.git.yu.c.chen@intel.com
Link: https://lkml.kernel.org/r/7ef90a88602ed536be46eba7152ed0d33bad5790.1748002400.git.yu.c.chen@intel.com
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Tested-by: K Prateek Nayak <kprateek.nayak@amd.com>
Tested-by: Madadi Vineeth Reddy <vineethr@linux.ibm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aubrey Li <aubrey.li@intel.com>
Cc: Ayush Jain <Ayush.jain3@amd.com>
Cc: "Chen, Tim C" <tim.c.chen@intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Libo Chen <libo.chen@oracle.com>
Cc: Mel Gorman <mgorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Michal Koutný <mkoutny@suse.com>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The current comment in gup_fast() talks about "IPIs that come from THPs
splitting", which is outdated and refers to the old THP splitting
implementation that was removed in commit ad0bed24e9 ("thp: drop all
split_huge_page()-related code"), which landed in v4.5. Before then, THP
splitting involved a pmdp_splitting_flush(), which sent an IPI to
serialize against gup_fast().
Nowadays, we use tlb_remove_table_sync_one() to send IPIs that serialize
against gup_fast(); this is used, for example, in THP *collapsing* to stop
gup_fast() walks of a page table before depositing it.
Link: https://lkml.kernel.org/r/20250528-gup-irq-comment-fix-v1-1-b9d83c345333@google.com
Signed-off-by: Jann Horn <jannh@google.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: John Hubbard <jhubbard@nvidia.com>
Cc: Kirill A. Shuemov <kirill.shutemov@linux.intel.com>
Cc: Peter Xu <peterx@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Problem: On large page size configurations (16KiB, 64KiB), the CMA
alignment requirement (CMA_MIN_ALIGNMENT_BYTES) increases considerably,
and this causes the CMA reservations to be larger than necessary. This
means that system will have less available MIGRATE_UNMOVABLE and
MIGRATE_RECLAIMABLE page blocks since MIGRATE_CMA can't fallback to them.
The CMA_MIN_ALIGNMENT_BYTES increases because it depends on MAX_PAGE_ORDER
which depends on ARCH_FORCE_MAX_ORDER. The value of ARCH_FORCE_MAX_ORDER
increases on 16k and 64k kernels.
For example, in ARM, the CMA alignment requirement when:
- CONFIG_ARCH_FORCE_MAX_ORDER default value is used
- CONFIG_TRANSPARENT_HUGEPAGE is set:
PAGE_SIZE | MAX_PAGE_ORDER | pageblock_order | CMA_MIN_ALIGNMENT_BYTES
-----------------------------------------------------------------------
4KiB | 10 | 9 | 4KiB * (2 ^ 9) = 2MiB
16Kib | 11 | 11 | 16KiB * (2 ^ 11) = 32MiB
64KiB | 13 | 13 | 64KiB * (2 ^ 13) = 512MiB
There are some extreme cases for the CMA alignment requirement when:
- CONFIG_ARCH_FORCE_MAX_ORDER maximum value is set
- CONFIG_TRANSPARENT_HUGEPAGE is NOT set:
- CONFIG_HUGETLB_PAGE is NOT set
PAGE_SIZE | MAX_PAGE_ORDER | pageblock_order | CMA_MIN_ALIGNMENT_BYTES
------------------------------------------------------------------------
4KiB | 15 | 15 | 4KiB * (2 ^ 15) = 128MiB
16Kib | 13 | 13 | 16KiB * (2 ^ 13) = 128MiB
64KiB | 13 | 13 | 64KiB * (2 ^ 13) = 512MiB
This affects the CMA reservations for the drivers. If a driver in a
4KiB kernel needs 4MiB of CMA memory, in a 16KiB kernel, the minimal
reservation has to be 32MiB due to the alignment requirements:
reserved-memory {
...
cma_test_reserve: cma_test_reserve {
compatible = "shared-dma-pool";
size = <0x0 0x400000>; /* 4 MiB */
...
};
};
reserved-memory {
...
cma_test_reserve: cma_test_reserve {
compatible = "shared-dma-pool";
size = <0x0 0x2000000>; /* 32 MiB */
...
};
};
Solution: Add a new config CONFIG_PAGE_BLOCK_ORDER that allows to set the
page block order in all the architectures. The maximum page block order
will be given by ARCH_FORCE_MAX_ORDER.
By default, CONFIG_PAGE_BLOCK_ORDER will have the same value that
ARCH_FORCE_MAX_ORDER. This will make sure that current kernel
configurations won't be affected by this change. It is a opt-in change.
This patch will allow to have the same CMA alignment requirements for
large page sizes (16KiB, 64KiB) as that in 4kb kernels by setting a lower
pageblock_order.
Tests:
- Verified that HugeTLB pages work when pageblock_order is 1, 7, 10 on
4k and 16k kernels.
- Verified that Transparent Huge Pages work when pageblock_order is 1,
7, 10 on 4k and 16k kernels.
- Verified that dma-buf heaps allocations work when pageblock_order is
1, 7, 10 on 4k and 16k kernels.
Benchmarks:
The benchmarks compare 16kb kernels with pageblock_order 10 and 7. The
reason for the pageblock_order 7 is because this value makes the min CMA
alignment requirement the same as that in 4kb kernels (2MB).
- Perform 100K dma-buf heaps (/dev/dma_heap/system) allocations of
SZ_8M, SZ_4M, SZ_2M, SZ_1M, SZ_64, SZ_8, SZ_4. Use simpleperf
(https://developer.android.com/ndk/guides/simpleperf) to measure the #
of instructions and page-faults on 16k kernels. The benchmark was
executed 10 times. The averages are below:
# instructions | #page-faults
order 10 | order 7 | order 10 | order 7
--------------------------------------------------------
13,891,765,770 | 11,425,777,314 | 220 | 217
14,456,293,487 | 12,660,819,302 | 224 | 219
13,924,261,018 | 13,243,970,736 | 217 | 221
13,910,886,504 | 13,845,519,630 | 217 | 221
14,388,071,190 | 13,498,583,098 | 223 | 224
13,656,442,167 | 12,915,831,681 | 216 | 218
13,300,268,343 | 12,930,484,776 | 222 | 218
13,625,470,223 | 14,234,092,777 | 219 | 218
13,508,964,965 | 13,432,689,094 | 225 | 219
13,368,950,667 | 13,683,587,37 | 219 | 225
-------------------------------------------------------------------
13,803,137,433 | 13,131,974,268 | 220 | 220 Averages
There were 4.85% #instructions when order was 7, in comparison with order
10.
13,803,137,433 - 13,131,974,268 = -671,163,166 (-4.86%)
The number of page faults in order 7 and 10 were the same.
These results didn't show any significant regression when the
pageblock_order is set to 7 on 16kb kernels.
- Run speedometer 3.1 (https://browserbench.org/Speedometer3.1/) 5 times
on the 16k kernels with pageblock_order 7 and 10.
order 10 | order 7 | order 7 - order 10 | (order 7 - order 10) %
-------------------------------------------------------------------
15.8 | 16.4 | 0.6 | 3.80%
16.4 | 16.2 | -0.2 | -1.22%
16.6 | 16.3 | -0.3 | -1.81%
16.8 | 16.3 | -0.5 | -2.98%
16.6 | 16.8 | 0.2 | 1.20%
-------------------------------------------------------------------
16.44 16.4 -0.04 -0.24% Averages
The results didn't show any significant regression when the
pageblock_order is set to 7 on 16kb kernels.
Link: https://lkml.kernel.org/r/20250521215807.1860663-1-jyescas@google.com
Signed-off-by: Juan Yescas <jyescas@google.com>
Acked-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Liam R. Howlett <Liam.Howlett@oracle.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Minchan Kim <minchan@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Commit b67fbebd4c ("mmu_gather: Force tlb-flush VM_PFNMAP vmas") added a
forced tlbflush to tlb_vma_end(), which is required to avoid a race
between munmap() and unmap_mapping_range(). However it added some
overhead to other paths where tlb_vma_end() is used, but vmas are not
removed, e.g. madvise(MADV_DONTNEED).
Fix this by moving the tlb flush out of tlb_end_vma() into new
tlb_flush_vmas() called from free_pgtables(), somewhat similar to the
stable version of the original commit: commit 895428ee124a ("mm: Force TLB
flush for PFNMAP mappings before unlink_file_vma()").
Note, that if tlb->fullmm is set, no flush is required, as the whole mm is
about to be destroyed.
Link: https://lkml.kernel.org/r/20250522012838.163876-1-roman.gushchin@linux.dev
Signed-off-by: Roman Gushchin <roman.gushchin@linux.dev>
Reviewed-by: Jann Horn <jannh@google.com>
Acked-by: Hugh Dickins <hughd@google.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Will Deacon <will@kernel.org>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@kernel.org>
Cc: Nick Piggin <npiggin@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
As of this writing, multiple major distros including Alma, Amazon,
Android, CentOS, Debian, Fedora, and Oracle are build-enabling DAMON (set
CONFIG_DAMON[1]). Enabling it by default will save configuration setup
time for the current and future DAMON users.
Build-enabling DAMON does not introduce a real risk since it makes no
behavioral change by default. It requires explicit user requests to do
anything. Only one potential risk is making the size of the kernel a
little bit larger. On a production-purpose configuration, it increases
the resulting kernel package size by about 0.1 % of the final package
file. I believe that's too small to be a real problem in common setups.
Hence, the benefit of enabling CONFIG_DAMON outweighs the potential risk.
Set CONFIG_DAMON by default.
Link: https://oracle.github.io/kconfigs/?config=UTS_RELEASE&config=DAMON [1]
Link: https://lkml.kernel.org/r/20250521042755.39653-3-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Acked-by: Honggyu Kim <honggyu.kim@sk.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "mm/damon: build-enable essential DAMON components by
default".
As of this writing, multiple major distros including Alma, Amazon,
Android, CentOS, Debian, Fedora, and Oracle are build-enabling DAMON (set
CONFIG_DAMON[1]). Configuring DAMON is not very easy, since it is
disabled by default, and there are multiple essential options that need to
be manually turned on, one by one. Make it easier, by grouping essential
configurations to be enabled with one selection, and enabling build of the
essential parts of DAMON by default.
Note that build-enabling DAMON does not introduce any real risk, since it
makes no behavioral change by default. It requires explicit user requests
to do anything. Only one potential risk is making the size of the kernel
a little bit larger. On a production-purpose configuration, it increases
the resulting kernel package binary size by about 0.1 % of the final
package file. I believe that's too small to be a real problem in common
setups.
DAMON_{VADDR,PADDR,SYSFS} are de-facto essential parts of DAMON for normal
usages. Because those need to be enabled one by one, however, and there
are other test-purpose or non-essential configurations, it is easy to be
confused and make mistakes at setup. Make the essential configurations
default to CONFIG_DAMON, so that those can be enabled by default with a
single change.
Link: https://oracle.github.io/kconfigs/?config=UTS_RELEASE&config=DAMON [1]
Link: https://lkml.kernel.org/r/20250521042755.39653-1-sj@kernel.org
Link: https://lkml.kernel.org/r/20250521042755.39653-2-sj@kernel.org
Signed-off-by: SeongJae Park <sj@kernel.org>
Acked-by: Honggyu Kim <honggyu.kim@sk.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
If multi shmem_unuse() for different swap type is called concurrently, a
dead loop could occur as following:
shmem_unuse(typeA) shmem_unuse(typeB)
mutex_lock(&shmem_swaplist_mutex)
list_for_each_entry_safe(info, next, ...)
...
mutex_unlock(&shmem_swaplist_mutex)
/* info->swapped may drop to 0 */
shmem_unuse_inode(&info->vfs_inode, type)
mutex_lock(&shmem_swaplist_mutex)
list_for_each_entry(info, next, ...)
if (!info->swapped)
list_del_init(&info->swaplist)
...
mutex_unlock(&shmem_swaplist_mutex)
mutex_lock(&shmem_swaplist_mutex)
/* iterate with offlist entry and encounter a dead loop */
next = list_next_entry(info, swaplist);
...
Restart the iteration if the inode is already off shmem_swaplist list to
fix the issue.
Link: https://lkml.kernel.org/r/20250516170939.965736-4-shikemeng@huaweicloud.com
Fixes: b56a2d8af9 ("mm: rid swapoff of quadratic complexity")
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Kairui Song <kasong@tencent.com>
Cc: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
When the number of the monitoring targets in running contexts is reduced,
there may be DAMOS quotas referencing the targets that will be destroyed.
Applying the scheme action for such DAMOS scheme will be skipped forever
looking for the starting part of the region for the destroyed monitoring
target.
To fix this issue, when the monitoring target is destroyed, reset the
starting part for all DAMOS quotas that reference the target.
Link: https://lkml.kernel.org/r/20250517141852.142802-1-akinobu.mita@gmail.com
Fixes: da87878010 ("mm/damon/sysfs: support online inputs update")
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Reviewed-by: SeongJae Park <sj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Patch series "memcg: nmi-safe kmem charging", v4.
Users can attached their BPF programs at arbitrary execution points in the
kernel and such BPF programs may run in nmi context. In addition, these
programs can trigger memcg charged kernel allocations in the nmi context.
However memcg charging infra for kernel memory is not equipped to handle
nmi context for all architectures.
This series removes the hurdles to enable kmem charging in the nmi context
for most of the archs. For archs without CONFIG_HAVE_NMI, this series is
a noop. For archs with NMI support and have
CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS, the previous work to make memcg
stats re-entrant is sufficient for allowing kmem charging in nmi context.
For archs with NMI support but without
CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS and with ARCH_HAVE_NMI_SAFE_CMPXCHG,
this series added infra to support kmem charging in nmi context. Lastly
those archs with NMI support but without
CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS and ARCH_HAVE_NMI_SAFE_CMPXCHG, kmem
charging in nmi context is not supported at all.
Mostly used archs have support for CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS
and this series should be almost a noop (other than making
memcg_rstat_updated nmi safe) for such archs.
This patch (of 5):
The memcg accounting and stats uses this_cpu* and atomic* ops. There are
archs which define CONFIG_HAVE_NMI but does not define
CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS and ARCH_HAVE_NMI_SAFE_CMPXCHG, so
memcg accounting for such archs in nmi context is not possible to support.
Let's just disable memcg accounting in nmi context for such archs.
Link: https://lkml.kernel.org/r/20250519063142.111219-1-shakeel.butt@linux.dev
Link: https://lkml.kernel.org/r/20250519063142.111219-2-shakeel.butt@linux.dev
Signed-off-by: Shakeel Butt <shakeel.butt@linux.dev>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
The mlock2-tests test_mlock_lock() test reports two test results with an
identical string, one reporitng if it successfully locked a block of
memory and another reporting if the lock is still present after doing an
unlock (following a similar pattern to other tests in the same program).
This confuses test automation since the test string is used to deduplicate
tests, change the post unlock test to report "Unlocked" instead like the
other tests to fix this.
Link: https://lkml.kernel.org/r/20250515-selftest-mm-mlock2-dup-v1-1-963d5d7d243a@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Acked-by: Dev Jain <dev.jain@arm.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
In old pcp design, pcp->free_factor gets incremented in nr_pcp_free()
which is invoked by free_pcppages_bulk(). So, it used to increase
free_factor by 1 only when we try to reduce the size of pcp list and
free_high used to trigger only for order > 0 and order < costly_order
and pcp->free_factor > 0.
For iperf3 I noticed that with older design in kernel v6.6, pcp list
was drained mostly when pcp->count > high (more often when count goes
above 530). and most of the time pcp->free_factor was 0, triggering
very few high order flushes.
But this is changed in the current design, introduced in commit
6ccdcb6d3a ("mm, pcp: reduce detecting time of consecutive high order
page freeing"), where pcp->free_factor is changed to pcp->free_count to
keep track of the number of pages freed contiguously. In this design,
pcp->free_count is incremented on every deallocation, irrespective of
whether pcp list was reduced or not. And logic to trigger free_high is
if pcp->free_count goes above batch (which is 63) and there are two
contiguous page free without any allocation.
With this design, for iperf3, pcp list is getting flushed more
frequently because free_high heuristics is triggered more often now. I
observed that high order pcp list is drained as soon as both count and
free_count goes above 63.
Due to this more aggressive high order flushing, applications doing
contiguous high order allocation will require to go to global list more
frequently.
On a 2-node AMD machine with 384 vCPUs on each node, connected via
Mellonox connectX-7, I am seeing a ~30% performance reduction if we
scale number of iperf3 client/server pairs from 32 to 64.
Though this new design reduced the time to detect high order flushes,
but for application which are allocating high order pages more
frequently it may be flushing the high order list pre-maturely. This
motivates towards tuning on how late or early we should flush high
order lists.
So, in this patch, we increased the pcp->free_count threshold to
trigger free_high from "batch" to "batch + pcp->high_min / 2" as
suggested by Ying [1], In the original pcp->free_factor solution,
free_high is triggered for contiguous freeing with size ranging from
"batch" to "pcp->high + batch". So, the average value is "batch +
pcp->high / 2". While in the pcp->free_count solution, free_high will
be triggered for contiguous freeing with size "batch". So, to restore
the original behavior, we can use the threshold "batch + pcp->high_min
/ 2"
This new threshold keeps high order pages in pcp list for a longer
duration which can help the application doing high order allocations
frequently.
With this patch performace to Iperf3 is restored and score for other
benchmarks on the same machine are as follows:
iperf3 lmbench3 netperf kbuild
(AF_UNIX) (SCTP_STREAM_MANY)
------- --------- ----------------- ------
v6.6 vanilla (base) 100 100 100 100
v6.12 vanilla 69 113 98.5 98.8
v6.12 + this patch 100 110.3 100.2 99.3
netperf-tcp:
6.12 6.12
vanilla this_patch
Hmean 64 732.14 ( 0.00%) 730.45 ( -0.23%)
Hmean 128 1417.46 ( 0.00%) 1419.44 ( 0.14%)
Hmean 256 2679.67 ( 0.00%) 2676.45 ( -0.12%)
Hmean 1024 8328.52 ( 0.00%) 8339.34 ( 0.13%)
Hmean 2048 12716.98 ( 0.00%) 12743.68 ( 0.21%)
Hmean 3312 15787.79 ( 0.00%) 15887.25 ( 0.63%)
Hmean 4096 17311.91 ( 0.00%) 17332.68 ( 0.12%)
Hmean 8192 20310.73 ( 0.00%) 20465.09 ( 0.76%)
Link: https://lore.kernel.org/all/875xjmuiup.fsf@DESKTOP-5N7EMDA/ [1]
Link: https://lkml.kernel.org/r/20250407105219.55351-1-nikhil.dhama@amd.com
Fixes: 6ccdcb6d3a ("mm, pcp: reduce detecting time of consecutive high order page freeing")
Signed-off-by: Nikhil Dhama <nikhil.dhama@amd.com>
Suggested-by: Huang Ying <ying.huang@linux.alibaba.com>
Reviewed-by: Huang Ying <ying.huang@linux.alibaba.com>
Cc: Raghavendra K T <raghavendra.kt@amd.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Bharata B Rao <bharata@amd.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
