]> git.proxmox.com Git - mirror_ubuntu-kernels.git/commit
mm/memcg: optimize memory.numa_stat like memory.stat
authorShakeel Butt <shakeelb@google.com>
Wed, 3 Jun 2020 22:56:24 +0000 (15:56 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 4 Jun 2020 03:09:42 +0000 (20:09 -0700)
commitdd8657b6c1cb5e65b13445b4a038736e81cf80ea
treeb09c47c6ec674a6b842b9ba4c1de1a04cd445344
parentdde3c6b72a16c2db826f54b2d49bdea26c3534a2
mm/memcg: optimize memory.numa_stat like memory.stat

Currently reading memory.numa_stat traverses the underlying memcg tree
multiple times to accumulate the stats to present the hierarchical view of
the memcg tree.  However the kernel already maintains the hierarchical
view of the stats and use it in memory.stat.  Just use the same mechanism
in memory.numa_stat as well.

I ran a simple benchmark which reads root_mem_cgroup's memory.numa_stat
file in the presense of 10000 memcgs.  The results are:

Without the patch:
$ time cat /dev/cgroup/memory/memory.numa_stat > /dev/null

real    0m0.700s
user    0m0.001s
sys     0m0.697s

With the patch:
$ time cat /dev/cgroup/memory/memory.numa_stat > /dev/null

real    0m0.001s
user    0m0.001s
sys     0m0.000s

[akpm@linux-foundation.org: avoid forcing out-of-line code generation]
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Michal Hocko <mhocko@kernel.org>
Link: http://lkml.kernel.org/r/20200304022058.248270-1-shakeelb@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/memcontrol.c