Thread (16 messages) 16 messages, 5 authors, 2012-11-23

Re: [PATCH 1/2] cgroup: helper do determine group name

From: Michal Hocko <hidden>
Date: 2012-11-23 08:55:38

On Thu 22-11-12 14:29:49, Glauber Costa wrote:
With more than one user, it is useful to have a helper function in the
cgroup core to derive a group's name.

We'll just return a pointer, and it is not expected to get incredibly
complicated. But it is useful to have it so we can abstract away the
vfs relation from its users.

Signed-off-by: Glauber Costa <redacted>
CC: Tejun Heo <tj@kernel.org>
CC: Michal Hocko <redacted>
CC: Kamezawa Hiroyuki <redacted>
CC: Johannes Weiner <hannes@cmpxchg.org>
Looks good to me in general. Minor comments bellow.
Anyway.
Acked-by: Michal Hocko <redacted>
---
Tejun:

I know the rcu is no longer necessary. I am using mhocko's tree,
that doesn't seem to have your last stream of patches yet.
Which patches are we talking about? Are they in a pullable (for -mm)
branch or I have to cherry-pick them?
quoted hunk ↗ jump to hunk
If you approve the interface, we'll need a follow up on this to remove
the rcu dereference of the dentry.

 include/linux/cgroup.h |  1 +
 kernel/cgroup.c        |  9 +++++++++
 mm/memcontrol.c        | 11 ++++-------
 3 files changed, 14 insertions(+), 7 deletions(-)
diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index a178a91..57c4ab1 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -401,6 +401,7 @@ int cgroup_rm_cftypes(struct cgroup_subsys *ss, const struct cftype *cfts);
 int cgroup_is_removed(const struct cgroup *cgrp);
 
 int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen);
+extern const char *cgroup_name(const struct cgroup *cgrp);
 
 int cgroup_task_count(const struct cgroup *cgrp);
 
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 3d68aad..d0d291e 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1757,6 +1757,15 @@ int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen)
 }
 EXPORT_SYMBOL_GPL(cgroup_path);
 
This expects css reference at caller, right. Please make it explicit
here in case somebody wants to use this somewhere else.
Besides that rcu_read_{un}lock are not necessary if you keep the
reference, right? The last dput happens only after the last css_put.
quoted hunk ↗ jump to hunk
+const char *cgroup_name(const struct cgroup *cgrp)
+{
+	struct dentry *dentry;
+	rcu_read_lock();
+	dentry = rcu_dereference_check(cgrp->dentry, cgroup_lock_is_held());
+	rcu_read_unlock();
+	return dentry->d_name.name;
+}
+
 /*
  * Control Group taskset
  */
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index e3d805f..05b87aa 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -3141,16 +3141,13 @@ void mem_cgroup_destroy_cache(struct kmem_cache *cachep)
 static char *memcg_cache_name(struct mem_cgroup *memcg, struct kmem_cache *s)
 {
 	char *name;
-	struct dentry *dentry;
+	const char *cgname;
 
-	rcu_read_lock();
-	dentry = rcu_dereference(memcg->css.cgroup->dentry);
-	rcu_read_unlock();
-
-	BUG_ON(dentry == NULL);
+	cgname = cgroup_name(memcg->css.cgroup);
+	BUG_ON(cgname == NULL);
 
 	name = kasprintf(GFP_KERNEL, "%s(%d:%s)", s->name,
-			 memcg_cache_id(memcg), dentry->d_name.name);
+			 memcg_cache_id(memcg), cgname);
 
 	return name;
 }
-- 
1.7.11.7

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help