[PATCH] cgroups: fix lock inconsistency in cgroup_clone()

From: Li Zefan
Date: Tue Jan 20 2009 - 01:45:22 EST


I fixed a bug in cgroup_clone() in Linus' tree (commit 7b574b7), without
noticing there was a cleanup patch in -mm tree that should be rebased
(now commit 104cbd5), thus resulted in lock inconsistency.

Signed-off-by: Li Zefan <lizf@xxxxxxxxxxxxxx>
---

for 2.6.29

---
kernel/cgroup.c | 13 +++++++------
1 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index c298310..a5826f3 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -2991,20 +2991,21 @@ int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys,
mutex_unlock(&cgroup_mutex);
return 0;
}
- task_lock(tsk);
- cg = tsk->cgroups;
- parent = task_cgroup(tsk, subsys->subsys_id);

/* Pin the hierarchy */
- if (!atomic_inc_not_zero(&parent->root->sb->s_active)) {
+ if (!atomic_inc_not_zero(&root->sb->s_active)) {
/* We race with the final deactivate_super() */
mutex_unlock(&cgroup_mutex);
return 0;
}

/* Keep the cgroup alive */
+ task_lock(tsk);
+ parent = task_cgroup(tsk, subsys->subsys_id);
+ cg = tsk->cgroups;
get_css_set(cg);
task_unlock(tsk);
+
mutex_unlock(&cgroup_mutex);

/* Now do the VFS work to create a cgroup */
@@ -3043,7 +3044,7 @@ int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys,
mutex_unlock(&inode->i_mutex);
put_css_set(cg);

- deactivate_super(parent->root->sb);
+ deactivate_super(root->sb);
/* The cgroup is still accessible in the VFS, but
* we're not going to try to rmdir() it at this
* point. */
@@ -3069,7 +3070,7 @@ int cgroup_clone(struct task_struct *tsk, struct cgroup_subsys *subsys,
mutex_lock(&cgroup_mutex);
put_css_set(cg);
mutex_unlock(&cgroup_mutex);
- deactivate_super(parent->root->sb);
+ deactivate_super(root->sb);
return ret;
}

--
1.5.4.rc3


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/