I think this is the right way to handle the lockdep false-positive in the
current cgroups patches, but I'm not that familiar with lockdep so any
suggestions for a better approach are welcomed.

In order to avoid a false-positive lockdep warning, we lock the root inode
of a new filesystem mount prior to taking cgroup_mutex, to preserve the
invariant that cgroup_mutex nests inside inode->i_mutex.  In order to
prevent a lockdep false positive when locking i_mutex on a newly-created
cgroup directory inode we use mutex_lock_nested(), with a nesting level
of I_MUTEX_CHILD since the new inode will ultimately be a child directory
of the parent whose i_mutex is nested outside of cgroup_mutex.

Signed-off-by: Paul Menage <menage@google.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
---

 kernel/cgroup.c |   17 +++++++----------
 1 files changed, 7 insertions(+), 10 deletions(-)

diff -puN kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays-avoid-lockdep-warning kernel/cgroup.c
--- a/kernel/cgroup.c~task-cgroupsv11-shared-cgroup-subsystem-group-arrays-avoid-lockdep-warning
+++ a/kernel/cgroup.c
@@ -867,13 +867,16 @@ static int cgroup_get_sb(struct file_
 	} else {
 		/* New superblock */
 		struct cgroup *cont = &root->top_cgroup;
+		struct inode *inode;
 
 		BUG_ON(sb->s_root != NULL);
 
 		ret = cgroup_get_rootdir(sb);
 		if (ret)
 			goto drop_new_super;
+		inode = sb->s_root->d_inode;
 
+		mutex_lock(&inode->i_mutex);
 		mutex_lock(&cgroup_mutex);
 
 		/*
@@ -886,12 +889,14 @@ static int cgroup_get_sb(struct file_
 		ret = allocate_cg_links(css_set_count, &tmp_cg_links);
 		if (ret) {
 			mutex_unlock(&cgroup_mutex);
+			mutex_unlock(&inode->i_mutex);
 			goto drop_new_super;
 		}
 
 		ret = rebind_subsystems(root, root->subsys_bits);
 		if (ret == -EBUSY) {
 			mutex_unlock(&cgroup_mutex);
+			mutex_unlock(&inode->i_mutex);
 			goto drop_new_super;
 		}
 
@@ -931,16 +936,8 @@ static int cgroup_get_sb(struct file_
 		BUG_ON(!list_empty(&cont->children));
 		BUG_ON(root->number_of_cgroups != 1);
 
-		/*
-		 * I believe that it's safe to nest i_mutex inside
-		 * cgroup_mutex in this case, since no-one else can
-		 * be accessing this directory yet. But we still need
-		 * to teach lockdep that this is the case - currently
-		 * a cgroupfs remount triggers a lockdep warning
-		 */
-		mutex_lock(&cont->dentry->d_inode->i_mutex);
 		cgroup_populate_dir(cont);
-		mutex_unlock(&cont->dentry->d_inode->i_mutex);
+		mutex_unlock(&inode->i_mutex);
 		mutex_unlock(&cgroup_mutex);
 	}
 
@@ -1358,7 +1355,7 @@ static int cgroup_create_file(struct 
 
 		/* start with the directory inode held, so that we can
 		 * populate it without racing with another mkdir */
-		mutex_lock(&inode->i_mutex);
+		mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
 	} else if (S_ISREG(mode)) {
 		inode->i_size = 0;
 		inode->i_fop = &cgroup_file_operations;
_

--
_______________________________________________
Containers mailing list
Containers@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers