Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1344417

Re: [PATCH v2 16/17] mm/slab: introduce new slab management type, OBJFREELIST_SLAB

Show all headers | View raw

On Fri, 26 Feb 2016, js1304@gmail.com wrote:

> Although this idea can apply to all caches whose size is larger than
> management array size, it isn't applied to caches which have a
> constructor.  If such cache's object is used for management array,
> constructor should be called for it before that object is returned to
> user.  I guess that overhead overwhelm benefit in that case so this idea
> doesn't applied to them at least now.

Caches which have a constructor (or are used with SLAB_RCU_FREE) have a
defined content even when they are free. Therefore they cannot be used
for the freelist.

> For summary, from now on, slab management type is determined by
> following logic.
>
> 1) if management array size is smaller than object size and no ctor, it
>    becomes OBJFREELIST_SLAB.

Also do not do this for RCU slabs.

> 2) if management array size is smaller than leftover, it becomes
>    NORMAL_SLAB which uses leftover as a array.
>
> 3) if OFF_SLAB help to save memory than way 4), it becomes OFF_SLAB.
>    It allocate a management array from the other cache so memory waste
>    happens.

Wonder how many of these ugly off slabs are left after what you did here.

> TOTAL = OBJFREELIST + NORMAL(leftover) + NORMAL + OFF
>
> /Before/
> 126 = 0 + 60 + 25 + 41
>
> /After/
> 126 = 97 + 12 + 15 + 2
>
> Result shows that number of caches that doesn't waste memory increase
> from 60 to 109.

Great results.

> v2: fix SLAB_DESTROTY_BY_RCU cache type handling

Ok how are they handled now? Do not see that dealt with in the patch.

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH v2 00/17] mm/slab: introduce new freed objects management way, OBJFREELIST_SLAB js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 16/17] mm/slab: introduce new slab management type, OBJFREELIST_SLAB js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 16/17] mm/slab: introduce new slab management type,  OBJFREELIST_SLAB Christoph Lameter <cl@linux.com> - 2016-02-26 17:30 +0100
      Re: [PATCH v2 16/17] mm/slab: introduce new slab management type, OBJFREELIST_SLAB Joonsoo Kim <js1304@gmail.com> - 2016-02-26 18:10 +0100
  [PATCH v2 11/17] mm/slab: clean up cache type determination js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 03/17] mm/slab: remove the checks for slab implementation bug js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 03/17] mm/slab: remove the checks for slab implementation  bug Christoph Lameter <cl@linux.com> - 2016-02-26 17:10 +0100
  [PATCH v2 01/17] mm/slab: fix stale code comment js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 04/17] mm/slab: activate debug_pagealloc in SLAB when it is actually enabled js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 04/17] mm/slab: activate debug_pagealloc in SLAB when  it is actually enabled Christoph Lameter <cl@linux.com> - 2016-02-26 17:10 +0100
  [PATCH v2 05/17] mm/slab: use more appropriate condition check for debug_pagealloc js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 05/17] mm/slab: use more appropriate condition check  for debug_pagealloc Christoph Lameter <cl@linux.com> - 2016-02-26 17:10 +0100
  [PATCH v2 09/17] mm/slab: put the freelist at the end of slab page js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 07/17] mm/slab: alternative implementation for DEBUG_SLAB_LEAK js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 13/17] mm/slab: make criteria for off slab determination robust and simple js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 12/17] mm/slab: do not change cache size if debug pagealloc isn't possible js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 12/17] mm/slab: do not change cache size if debug  pagealloc isn't possible Christoph Lameter <cl@linux.com> - 2016-02-26 17:20 +0100
      Re: [PATCH v2 12/17] mm/slab: do not change cache size if debug  pagealloc isn't possible Joonsoo Kim <js1304@gmail.com> - 2016-02-26 18:10 +0100
  [PATCH v2 17/17] mm/slab: avoid returning values by reference js1304@gmail.com - 2016-02-26 07:10 +0100
    Re: [PATCH v2 17/17] mm/slab: avoid returning values by reference Christoph Lameter <cl@linux.com> - 2016-02-26 17:30 +0100
  [PATCH v2 06/17] mm/slab: clean up DEBUG_PAGEALLOC processing code js1304@gmail.com - 2016-02-26 07:10 +0100
  [PATCH v2 08/17] mm/slab: remove object status buffer for DEBUG_SLAB_LEAK js1304@gmail.com - 2016-02-26 07:10 +0100

csiph-web