Commit 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
enabled memcg accounting for pids allocated from init_pid_ns.pid_cachep,
but forgot to adjust the setting for nested pid namespaces.
As a result, pid memory is not accounted exactly where it is really needed,
inside memcg-limited containers with their own pid namespaces.
Pid was one the first kernel objects enabled for memcg accounting.
init_pid_ns.pid_cachep marked by SLAB_ACCOUNT and we can expect that
any new pids in the system are memcg-accounted.
Though recently I've noticed that it is wrong. nested pid namespaces creates
own slab caches for pid objects, nested pids have increased size because contain
id both for all parent and for own pid namespaces. The problem is that these slab
caches are _NOT_ marked by SLAB_ACCOUNT, as a result any pids allocated in
nested pid namespaces are not memcg-accounted.
Pid struct in nested pid namespace consumes up to 500 bytes memory,
100000 such objects gives us up to ~50Mb unaccounted memory,
this allow container to exceed assigned memcg limits.
Fixes: 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
Cc: [email protected]
Signed-off-by: Vasily Averin <[email protected]>
Reviewed-by: Michal Koutný <[email protected]>
Acked-by: Christian Brauner <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
---
kernel/pid_namespace.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c
index 6cd6715..a46a372 100644
--- a/kernel/pid_namespace.c
+++ b/kernel/pid_namespace.c
@@ -51,7 +51,8 @@ static struct kmem_cache *create_pid_cachep(unsigned int level)
mutex_lock(&pid_caches_mutex);
/* Name collision forces to do allocation under mutex. */
if (!*pkc)
- *pkc = kmem_cache_create(name, len, 0, SLAB_HWCACHE_ALIGN, 0);
+ *pkc = kmem_cache_create(name, len, 0,
+ SLAB_HWCACHE_ALIGN | SLAB_ACCOUNT, 0);
mutex_unlock(&pid_caches_mutex);
/* current can fail, but someone else can succeed. */
return READ_ONCE(*pkc);
--
1.8.3.1
On Sat, Apr 24, 2021 at 4:54 AM Vasily Averin <[email protected]> wrote:
>
> Commit 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
> enabled memcg accounting for pids allocated from init_pid_ns.pid_cachep,
> but forgot to adjust the setting for nested pid namespaces.
> As a result, pid memory is not accounted exactly where it is really needed,
> inside memcg-limited containers with their own pid namespaces.
>
> Pid was one the first kernel objects enabled for memcg accounting.
> init_pid_ns.pid_cachep marked by SLAB_ACCOUNT and we can expect that
> any new pids in the system are memcg-accounted.
>
> Though recently I've noticed that it is wrong. nested pid namespaces creates
> own slab caches for pid objects, nested pids have increased size because contain
> id both for all parent and for own pid namespaces. The problem is that these slab
> caches are _NOT_ marked by SLAB_ACCOUNT, as a result any pids allocated in
> nested pid namespaces are not memcg-accounted.
>
> Pid struct in nested pid namespace consumes up to 500 bytes memory,
> 100000 such objects gives us up to ~50Mb unaccounted memory,
> this allow container to exceed assigned memcg limits.
>
> Fixes: 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
> Cc: [email protected]
> Signed-off-by: Vasily Averin <[email protected]>
> Reviewed-by: Michal Koutný <[email protected]>
> Acked-by: Christian Brauner <[email protected]>
> Acked-by: Roman Gushchin <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Dear Andrew,
could you please pick up this patch and add
Reviewed-by: Shakeel Butt <[email protected]>
Thank you,
Vasily Averin
On 4/24/21 2:54 PM, Vasily Averin wrote:
> Commit 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
> enabled memcg accounting for pids allocated from init_pid_ns.pid_cachep,
> but forgot to adjust the setting for nested pid namespaces.
> As a result, pid memory is not accounted exactly where it is really needed,
> inside memcg-limited containers with their own pid namespaces.
>
> Pid was one the first kernel objects enabled for memcg accounting.
> init_pid_ns.pid_cachep marked by SLAB_ACCOUNT and we can expect that
> any new pids in the system are memcg-accounted.
>
> Though recently I've noticed that it is wrong. nested pid namespaces creates
> own slab caches for pid objects, nested pids have increased size because contain
> id both for all parent and for own pid namespaces. The problem is that these slab
> caches are _NOT_ marked by SLAB_ACCOUNT, as a result any pids allocated in
> nested pid namespaces are not memcg-accounted.
>
> Pid struct in nested pid namespace consumes up to 500 bytes memory,
> 100000 such objects gives us up to ~50Mb unaccounted memory,
> this allow container to exceed assigned memcg limits.
>
> Fixes: 5d097056c9a0 ("kmemcg: account certain kmem allocations to memcg")
> Cc: [email protected]
> Signed-off-by: Vasily Averin <[email protected]>
> Reviewed-by: Michal Koutný <[email protected]>
> Acked-by: Christian Brauner <[email protected]>
> Acked-by: Roman Gushchin <[email protected]>
> ---
> kernel/pid_namespace.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c
> index 6cd6715..a46a372 100644
> --- a/kernel/pid_namespace.c
> +++ b/kernel/pid_namespace.c
> @@ -51,7 +51,8 @@ static struct kmem_cache *create_pid_cachep(unsigned int level)
> mutex_lock(&pid_caches_mutex);
> /* Name collision forces to do allocation under mutex. */
> if (!*pkc)
> - *pkc = kmem_cache_create(name, len, 0, SLAB_HWCACHE_ALIGN, 0);
> + *pkc = kmem_cache_create(name, len, 0,
> + SLAB_HWCACHE_ALIGN | SLAB_ACCOUNT, 0);
> mutex_unlock(&pid_caches_mutex);
> /* current can fail, but someone else can succeed. */
> return READ_ONCE(*pkc);
>