* [PATCH 2/2] memcg: fix kmem over-charging for embedded obj_exts array
[not found] <20260310113804.245647-1-ranxiaokai627@163.com>
@ 2026-03-10 11:38 ` ranxiaokai627
0 siblings, 0 replies; only message in thread
From: ranxiaokai627 @ 2026-03-10 11:38 UTC (permalink / raw)
To: hannes, mhocko, roman.gushchin, shakeel.butt, muchun.song, akpm,
vbabka, harry.yoo, hao.li, cl, rientjes
Cc: cgroups, linux-mm, linux-kernel, ran.xiaokai, ranxiaokai627,
stable
From: Ran Xiaokai <ran.xiaokai@zte.com.cn>
Since commit a77d6d338685 ("mm/slab: place slabobj_ext metadata
in unused space within s->size"), the struct slabobj_ext array
can use slab leftover space or be embedded into the slub object
to save memory by calling alloc_slab_obj_exts_early(). In these
cases, no extra kmalloc space is allocated to store the obj_exts array.
Optimize obj_full_size() behavior to avoid over-charging kmem
for objects whose obj_exts use slab leftover space or are embedded
in the object.
Fixes: a77d6d338685 ("mm/slab: place slabobj_ext metadata in unused space within s->size")
Cc: stable@vger.kernel.org
Signed-off-by: Ran Xiaokai <ran.xiaokai@zte.com.cn>
---
mm/memcontrol.c | 19 ++++++++++++++-----
1 file changed, 14 insertions(+), 5 deletions(-)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 87614cfc4a3e..d6289a5cd6f3 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -3199,11 +3199,20 @@ void obj_cgroup_uncharge(struct obj_cgroup *objcg, size_t size)
refill_obj_stock(objcg, size, true, 0, NULL, 0);
}
-static inline size_t obj_full_size(struct kmem_cache *s)
+static inline size_t obj_full_size(struct kmem_cache *s, struct slab *slab)
{
+ if (obj_exts_in_slab(s, slab)) {
+ /*
+ * If obj_exts array uses slab leftover space or is embedded
+ * in object, no extra space is allocated, just charge the
+ * object size.
+ */
+ return s->size;
+ }
+
/*
- * For each accounted object there is an extra space which is used
- * to store obj_cgroup membership. Charge it too.
+ * For caches whose obj_exts array is allocated separately
+ * outside of slab, also charge the objcg pointer.
*/
return s->size + sizeof(struct obj_cgroup *);
}
@@ -3270,7 +3279,7 @@ bool __memcg_slab_post_alloc_hook(struct kmem_cache *s, struct list_lru *lru,
* TODO: we could batch this until slab_pgdat(slab) changes
* between iterations, with a more complicated undo
*/
- if (obj_cgroup_charge_account(objcg, flags, obj_full_size(s),
+ if (obj_cgroup_charge_account(objcg, flags, obj_full_size(s, slab),
slab_pgdat(slab), cache_vmstat_idx(s)))
return false;
@@ -3289,7 +3298,7 @@ bool __memcg_slab_post_alloc_hook(struct kmem_cache *s, struct list_lru *lru,
void __memcg_slab_free_hook(struct kmem_cache *s, struct slab *slab,
void **p, int objects, unsigned long obj_exts)
{
- size_t obj_size = obj_full_size(s);
+ size_t obj_size = obj_full_size(s, slab);
for (int i = 0; i < objects; i++) {
struct obj_cgroup *objcg;
--
2.25.1
^ permalink raw reply related [flat|nested] only message in thread
only message in thread, other threads:[~2026-03-10 11:39 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
[not found] <20260310113804.245647-1-ranxiaokai627@163.com>
2026-03-10 11:38 ` [PATCH 2/2] memcg: fix kmem over-charging for embedded obj_exts array ranxiaokai627
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox