Message-ID: <20251027122847.320924-1-harry.yoo@oracle.com>
Date: Mon, 27 Oct 2025 21:28:40 +0900
From: Harry Yoo <harry.yoo@...cle.com>
To: akpm@...ux-foundation.org, vbabka@...e.cz
Cc: andreyknvl@...il.com, cl@...ux.com, dvyukov@...gle.com, glider@...gle.com,
        hannes@...xchg.org, linux-mm@...ck.org, mhocko@...nel.org,
        muchun.song@...ux.dev, rientjes@...gle.com, roman.gushchin@...ux.dev,
        ryabinin.a.a@...il.com, shakeel.butt@...ux.dev, surenb@...gle.com,
        vincenzo.frascino@....com, yeoreum.yun@....com, harry.yoo@...cle.com,
        tytso@....edu, adilger.kernel@...ger.ca, linux-ext4@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: [RFC PATCH V3 0/7] mm/slab: reduce slab accounting memory overhead by allocating slabobj_ext metadata within unused slab space 

RFC v2: https://lore.kernel.org/linux-mm/20250827113726.707801-1-harry.yoo@oracle.com/

RFC v2 -> v3:
  - RFC v3 now depends on the patch "[PATCH V2] mm/slab: ensure all metadata
    in slab object are word-aligned"

  - During the merge window, the size of ext4 inode cache has shrunken
    and it couldn't benefit from the change anymore as the unused space
    became smaller. But I somehow found a way to shrink the size of
    ext4 inode object by a word...

    With new patch 1 and 2, now it can benefit from the optimization again.

  - As suggested by Andrey, SLUB now disables KASAN and KMSAN, and reset the
    kasan tag instead of unpoisoning slabobj_ext metadata (Patch 5).

When CONFIG_MEMCG and CONFIG_MEM_ALLOC_PROFILING are enabled,
the kernel allocates two pointers per object: one for the memory cgroup
(obj_cgroup) to which it belongs, and another for the code location
that requested the allocation.

In two special cases, this overhead can be eliminated by allocating
slabobj_ext metadata from unused space within a slab:

  Case 1. The "leftover" space after the last slab object is larger than
          the size of an array of slabobj_ext.

  Case 2. The per-object alignment padding is larger than
          sizeof(struct slabobj_ext).

For these two cases, one or two pointers can be saved per slab object.
Examples: ext4 inode cache (case 1) and xfs inode cache (case 2).
That's approximately 0.7-0.8% (memcg) or 1.5-1.6%% (memcg + mem profiling)
of the total inode cache size.

Implementing case 2 is not straightforward, because the existing code
assumes that slab->obj_exts is an array of slabobj_ext, while case 2
breaks the assumption.

As suggested by Vlastimil, abstract access to individual slabobj_ext
metadata via a new helper named slab_obj_ext():

static inline struct slabobj_ext *slab_obj_ext(struct slab *slab,
                                               unsigned long obj_exts,
                                               unsigned int index)
{
        return (struct slabobj_ext *)(obj_exts + slab_get_stride(slab) * index);
} 

In the normal case (including case 1), slab->obj_exts points to an array
of slabobj_ext, and the stride is sizeof(struct slabobj_ext).

In case 2, the stride is s->size and
slab->obj_exts = slab_address(slab) + s->red_left_pad + (offset of slabobj_ext)

With this approach, the memcg charging fastpath doesn't need to care the
storage method of slabobj_ext.

Harry Yoo (7):
  mm/slab: allow specifying freepointer offset when using constructor
  ext4: specify the free pointer offset for ext4_inode_cache
  mm/slab: abstract slabobj_ext access via new slab_obj_ext() helper
  mm/slab: use stride to access slabobj_ext
  mm/memcontrol,alloc_tag: handle slabobj_ext access under KASAN poison
  mm/slab: save memory by allocating slabobj_ext array from leftover
  mm/slab: place slabobj_ext metadata in unused space within s->size

 fs/ext4/super.c      |  20 ++-
 include/linux/slab.h |   9 ++
 mm/memcontrol.c      |  34 +++--
 mm/slab.h            |  94 ++++++++++++-
 mm/slab_common.c     |   8 +-
 mm/slub.c            | 304 ++++++++++++++++++++++++++++++++++++-------
 6 files changed, 398 insertions(+), 71 deletions(-)

-- 
2.43.0

Powered by blists - more mailing lists