The quilt patch titled
Subject: memcg: rename do_flush_stats and add force flag
has been removed from the -mm tree. Its filename was
memcg-rename-do_flush_stats-and-add-force-flag.patch
This patch was dropped because it was merged into the mm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: JP Kobryn <inwardvessel@xxxxxxxxx>
Subject: memcg: rename do_flush_stats and add force flag
Date: Mon, 28 Oct 2024 19:11:05 -0700
Patch series "memcg: tracepoint for flushing stats", v3.
This series adds new capability for understanding frequency and circumstances
behind flushing memcg stats.
This patch (of 2):
Change the name to something more consistent with others in the file and
use double unders to signify it is associated with the
mem_cgroup_flush_stats() API call. Additionally include a new flag that
call sites use to indicate a forced flush; skipping checks and flushing
unconditionally. There are no changes in functionality.
Link: https://lkml.kernel.org/r/20241029021106.25587-1-inwardvessel@xxxxxxxxx
Link: https://lkml.kernel.org/r/20241029021106.25587-2-inwardvessel@xxxxxxxxx
Signed-off-by: JP Kobryn <inwardvessel@xxxxxxxxx>
Reviewed-by: Yosry Ahmed <yosryahmed@xxxxxxxxxx>
Acked-by: Shakeel Butt <shakeel.butt@xxxxxxxxx>
Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Steven Rostedt (Google) <rostedt@xxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/memcontrol.c | 17 ++++++++---------
1 file changed, 8 insertions(+), 9 deletions(-)
--- a/mm/memcontrol.c~memcg-rename-do_flush_stats-and-add-force-flag
+++ a/mm/memcontrol.c
@@ -595,8 +595,11 @@ static inline void memcg_rstat_updated(s
}
}
-static void do_flush_stats(struct mem_cgroup *memcg)
+static void __mem_cgroup_flush_stats(struct mem_cgroup *memcg, bool force)
{
+ if (!force && !memcg_vmstats_needs_flush(memcg->vmstats))
+ return;
+
if (mem_cgroup_is_root(memcg))
WRITE_ONCE(flush_last_time, jiffies_64);
@@ -620,8 +623,7 @@ void mem_cgroup_flush_stats(struct mem_c
if (!memcg)
memcg = root_mem_cgroup;
- if (memcg_vmstats_needs_flush(memcg->vmstats))
- do_flush_stats(memcg);
+ __mem_cgroup_flush_stats(memcg, false);
}
void mem_cgroup_flush_stats_ratelimited(struct mem_cgroup *memcg)
@@ -637,7 +639,7 @@ static void flush_memcg_stats_dwork(stru
* Deliberately ignore memcg_vmstats_needs_flush() here so that flushing
* in latency-sensitive paths is as cheap as possible.
*/
- do_flush_stats(root_mem_cgroup);
+ __mem_cgroup_flush_stats(root_mem_cgroup, true);
queue_delayed_work(system_unbound_wq, &stats_flush_dwork, FLUSH_TIME);
}
@@ -5286,11 +5288,8 @@ bool obj_cgroup_may_zswap(struct obj_cgr
break;
}
- /*
- * mem_cgroup_flush_stats() ignores small changes. Use
- * do_flush_stats() directly to get accurate stats for charging.
- */
- do_flush_stats(memcg);
+ /* Force flush to get accurate stats for charging */
+ __mem_cgroup_flush_stats(memcg, true);
pages = memcg_page_state(memcg, MEMCG_ZSWAP_B) / PAGE_SIZE;
if (pages < max)
continue;
_
Patches currently in -mm which might be from inwardvessel@xxxxxxxxx are
