[00/20] remove the SLAB allocator

Message

Vlastimil Babka Nov. 13, 2023, 7:13 p.m. UTC

The SLAB allocator has been deprecated since 6.5 and nobody has objected
so far. As we agreed at LSF/MM, we should wait with the removal until
the next LTS kernel is released. AFAIK that version hasn't been
announced yet, but assuming it would be 6.7, we can aim for 6.8 and
start exposing the removal to linux-next during the 6.7 cycle.

To keep the series reasonably sized and not pull in people from other
subsystems than mm and closely related ones, I didn't attempt to remove
every trace of unnecessary reference to dead config options in external
areas, nor in the defconfigs. Such cleanups can be sent to and handled
by respective maintainers after this is merged.

Instead I have added some patches aimed to reap some immediate benefits
of the removal, mainly by not having to split some fastpath code between
slab_common.c and slub.c anymore. But that is also not an exhaustive
effort and I expect more cleanups and optimizations will follow later.

Patch 08 updates CREDITS for the removed mm/slab.c. Please point out if
I missed someone not yet credited.

Git version: https://git.kernel.org/vbabka/l/slab-remove-slab-v1r4

Vlastimil Babka (20):
  mm/slab: remove CONFIG_SLAB from all Kconfig and Makefile
  KASAN: remove code paths guarded by CONFIG_SLAB
  KFENCE: cleanup kfence_guarded_alloc() after CONFIG_SLAB removal
  mm/memcontrol: remove CONFIG_SLAB #ifdef guards
  cpu/hotplug: remove CPUHP_SLAB_PREPARE hooks
  mm/slab: remove CONFIG_SLAB code from slab common code
  mm/mempool/dmapool: remove CONFIG_DEBUG_SLAB ifdefs
  mm/slab: remove mm/slab.c and slab_def.h
  mm/slab: move struct kmem_cache_cpu declaration to slub.c
  mm/slab: move the rest of slub_def.h to mm/slab.h
  mm/slab: consolidate includes in the internal mm/slab.h
  mm/slab: move pre/post-alloc hooks from slab.h to slub.c
  mm/slab: move memcg related functions from slab.h to slub.c
  mm/slab: move struct kmem_cache_node from slab.h to slub.c
  mm/slab: move kfree() from slab_common.c to slub.c
  mm/slab: move kmalloc_slab() to mm/slab.h
  mm/slab: move kmalloc() functions from slab_common.c to slub.c
  mm/slub: remove slab_alloc() and __kmem_cache_alloc_lru() wrappers
  mm/slub: optimize alloc fastpath code layout
  mm/slub: optimize free fast path code layout

 CREDITS                  |   12 +-
 arch/arm64/Kconfig       |    2 +-
 arch/s390/Kconfig        |    2 +-
 arch/x86/Kconfig         |    2 +-
 include/linux/slab.h     |   21 +-
 include/linux/slab_def.h |  124 --
 include/linux/slub_def.h |  204 --
 kernel/cpu.c             |    5 -
 lib/Kconfig.debug        |    1 -
 lib/Kconfig.kasan        |   11 +-
 lib/Kconfig.kfence       |    2 +-
 lib/Kconfig.kmsan        |    2 +-
 mm/Kconfig               |   50 +-
 mm/Kconfig.debug         |   16 +-
 mm/Makefile              |    6 +-
 mm/dmapool.c             |    2 +-
 mm/kasan/common.c        |   13 +-
 mm/kasan/kasan.h         |    3 +-
 mm/kasan/quarantine.c    |    7 -
 mm/kasan/report.c        |    1 +
 mm/kfence/core.c         |    4 -
 mm/memcontrol.c          |    6 +-
 mm/mempool.c             |    6 +-
 mm/slab.c                | 4026 --------------------------------------
 mm/slab.h                |  550 ++----
 mm/slab_common.c         |  231 +--
 mm/slub.c                |  597 +++++-
 27 files changed, 784 insertions(+), 5122 deletions(-)
 delete mode 100644 include/linux/slab_def.h
 delete mode 100644 include/linux/slub_def.h
 delete mode 100644 mm/slab.c

Comments

Michal Hocko Nov. 14, 2023, 11:18 a.m. UTC | #1

On Mon 13-11-23 20:13:41, Vlastimil Babka wrote:
> The SLAB allocator has been deprecated since 6.5 and nobody has objected
> so far. As we agreed at LSF/MM, we should wait with the removal until
> the next LTS kernel is released. AFAIK that version hasn't been
> announced yet, but assuming it would be 6.7, we can aim for 6.8 and
> start exposing the removal to linux-next during the 6.7 cycle.

Makes sense to me.
[...]
>  27 files changed, 784 insertions(+), 5122 deletions(-)

This is just too much of a maintenance burden to have a comfort of
multiple low level allocators. So it is good to see it go. Not that I
would have anything against SLAB allocator as such but if we need to
choose SLUB seems like a better choice.

Thanks for all the work!