On Tue, 28 Jan 2020, Michal Koutný wrote:
> On Mon, Jan 27, 2020 at 11:04:53PM +0000, Christopher Lameter <cl@xxxxxxxxx> wrote:
> > The patch exposes details of cgroup caches? Which patch are we talking
> > about?
> Sorry, that's misunderstanding. I mean the current state (sending
> uevents) exposes the internals (creation of caches per cgroup). The
> patch [1] removing uevent notifications is rectifying it.
From: Christoph Lameter <cl@xxxxxxxxx>
Subject: slub: Remove userspace notifier for cache add/remove
Kmem caches are internal kernel structures so it is strange that
userspace notifiers would be needed. And I am not aware of any use
of these notifiers. These notifiers may just exist because in the
initial slub release the sysfs code was copied from another
subsystem.
Signed-off-by: Christoph Lameter <cl@xxxxxxxxx>
Index: linux/mm/slub.c
===================================================================
--- linux.orig/mm/slub.c 2020-01-28 18:13:02.134506141 +0000
+++ linux/mm/slub.c 2020-01-28 18:13:02.134506141 +0000
@@ -5632,19 +5632,6 @@ static struct kobj_type slab_ktype = {
.release = kmem_cache_release,
};
-static int uevent_filter(struct kset *kset, struct kobject *kobj)
-{
- struct kobj_type *ktype = get_ktype(kobj);
-
- if (ktype == &slab_ktype)
- return 1;
- return 0;
-}
-
-static const struct kset_uevent_ops slab_uevent_ops = {
- .filter = uevent_filter,
-};
-
static struct kset *slab_kset;
static inline struct kset *cache_kset(struct kmem_cache *s)
@@ -5712,7 +5699,6 @@ static void sysfs_slab_remove_workfn(str
#ifdef CONFIG_MEMCG
kset_unregister(s->memcg_kset);
#endif
- kobject_uevent(&s->kobj, KOBJ_REMOVE);
out:
kobject_put(&s->kobj);
}
@@ -5770,7 +5756,6 @@ static int sysfs_slab_add(struct kmem_ca
}
#endif
- kobject_uevent(&s->kobj, KOBJ_ADD);
if (!unmergeable) {
/* Setup first alias */
sysfs_slab_alias(s, s->name);
@@ -5851,7 +5836,7 @@ static int __init slab_sysfs_init(void)
mutex_lock(&slab_mutex);
- slab_kset = kset_create_and_add("slab", &slab_uevent_ops, kernel_kobj);
+ slab_kset = kset_create_and_add("slab", NULL, kernel_kobj);
if (!slab_kset) {
mutex_unlock(&slab_mutex);
pr_err("Cannot register slab subsystem.\n");
