The patch titled
Subject: workingset: memcg: sleep when flushing stats in workingset_refault()
has been added to the -mm mm-unstable branch. Its filename is
workingset-memcg-sleep-when-flushing-stats-in-workingset_refault.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/workingset-memcg-sleep-when-flushing-stats-in-workingset_refault.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Yosry Ahmed <yosryahmed@xxxxxxxxxx>
Subject: workingset: memcg: sleep when flushing stats in workingset_refault()
Date: Thu, 30 Mar 2023 19:17:59 +0000
In workingset_refault(), we call
mem_cgroup_flush_stats_atomic_ratelimited() to read accurate stats within
an RCU read section and with sleeping disallowed. Move the call above the
RCU read section to make it non-atomic.
Flushing is an expensive operation that scales with the number of cpus and
the number of cgroups in the system, so avoid doing it atomically where
possible.
Since workingset_refault() is the only caller of
mem_cgroup_flush_stats_atomic_ratelimited(), just make it non-atomic, and
rename it to mem_cgroup_flush_stats_ratelimited().
Link: https://lkml.kernel.org/r/20230330191801.1967435-7-yosryahmed@xxxxxxxxxx
Signed-off-by: Yosry Ahmed <yosryahmed@xxxxxxxxxx>
Acked-by: Shakeel Butt <shakeelb@xxxxxxxxxx>
Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
Cc: Jens Axboe <axboe@xxxxxxxxx>
Cc: Josef Bacik <josef@xxxxxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Michal Koutný <mkoutny@xxxxxxxx>
Cc: Muchun Song <muchun.song@xxxxxxxxx>
Cc: Roman Gushchin <roman.gushchin@xxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: Vasily Averin <vasily.averin@xxxxxxxxx>
Cc: Zefan Li <lizefan.x@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/memcontrol.h | 4 ++--
mm/memcontrol.c | 4 ++--
mm/workingset.c | 5 +++--
3 files changed, 7 insertions(+), 6 deletions(-)
--- a/include/linux/memcontrol.h~workingset-memcg-sleep-when-flushing-stats-in-workingset_refault
+++ a/include/linux/memcontrol.h
@@ -1039,7 +1039,7 @@ static inline unsigned long lruvec_page_
void mem_cgroup_flush_stats(void);
void mem_cgroup_flush_stats_atomic(void);
-void mem_cgroup_flush_stats_atomic_ratelimited(void);
+void mem_cgroup_flush_stats_ratelimited(void);
void __mod_memcg_lruvec_state(struct lruvec *lruvec, enum node_stat_item idx,
int val);
@@ -1541,7 +1541,7 @@ static inline void mem_cgroup_flush_stat
{
}
-static inline void mem_cgroup_flush_stats_atomic_ratelimited(void)
+static inline void mem_cgroup_flush_stats_ratelimited(void)
{
}
--- a/mm/memcontrol.c~workingset-memcg-sleep-when-flushing-stats-in-workingset_refault
+++ a/mm/memcontrol.c
@@ -674,10 +674,10 @@ void mem_cgroup_flush_stats_atomic(void)
do_flush_stats(true);
}
-void mem_cgroup_flush_stats_atomic_ratelimited(void)
+void mem_cgroup_flush_stats_ratelimited(void)
{
if (time_after64(jiffies_64, READ_ONCE(flush_next_time)))
- mem_cgroup_flush_stats_atomic();
+ mem_cgroup_flush_stats();
}
static void flush_memcg_stats_dwork(struct work_struct *w)
--- a/mm/workingset.c~workingset-memcg-sleep-when-flushing-stats-in-workingset_refault
+++ a/mm/workingset.c
@@ -406,6 +406,9 @@ void workingset_refault(struct folio *fo
unpack_shadow(shadow, &memcgid, &pgdat, &eviction, &workingset);
eviction <<= bucket_order;
+ /* Flush stats (and potentially sleep) before holding RCU read lock */
+ mem_cgroup_flush_stats_ratelimited();
+
rcu_read_lock();
/*
* Look up the memcg associated with the stored ID. It might
@@ -461,8 +464,6 @@ void workingset_refault(struct folio *fo
lruvec = mem_cgroup_lruvec(memcg, pgdat);
mod_lruvec_state(lruvec, WORKINGSET_REFAULT_BASE + file, nr);
-
- mem_cgroup_flush_stats_atomic_ratelimited();
/*
* Compare the distance to the existing workingset size. We
* don't activate pages that couldn't stay resident even if
_
Patches currently in -mm which might be from yosryahmed@xxxxxxxxxx are
cgroup-rename-cgroup_rstat_flush_irqsafe-to-atomic.patch
memcg-rename-mem_cgroup_flush_stats_delayed-to-ratelimited.patch
memcg-do-not-flush-stats-in-irq-context.patch
memcg-replace-stats_flush_lock-with-an-atomic.patch
memcg-sleep-during-flushing-stats-in-safe-contexts.patch
workingset-memcg-sleep-when-flushing-stats-in-workingset_refault.patch
vmscan-memcg-sleep-when-flushing-stats-during-reclaim.patch
memcg-do-not-modify-rstat-tree-for-zero-updates.patch
