cgroup_clear_directroy is called by cgroup_d_remove_dir
and cgroup_remount.
when we call cgroup_remount to remount the cgroup,the subsystem
may be unlinked from cgroupfs_root->subsys_list in rebind_subsystem,this
subsystem's files will not be removed in cgroup_clear_directroy.
And the system will panic when we try to access these files.
this patch removes subsystems's files before rebind_subsystems,
if rebind_subsystems failed, repopulate these removed files.
With help from Tejun.
Signed-off-by: Gao feng <gaofeng@xxxxxxxxxxxxxx>
---
kernel/cgroup.c | 11 +++++++++--
1 files changed, 9 insertions(+), 2 deletions(-)
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index e1293a9..afabd2a 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1349,14 +1349,21 @@ static int cgroup_remount(struct super_block *sb, int *flags, char *data)
goto out_unlock;
}
+ /*
+ * Clear out the files of subsystems that should be removed,do
+ * this job before rebind_subsystems,since rebind_subsystems may
+ * change this hierarchy's subsys_list.
+ */
+ cgroup_clear_directory(cgrp->dentry, false, removed_mask);
+
ret = rebind_subsystems(root, opts.subsys_mask);
if (ret) {
+ /* rebind_subsystems failed,re-populate the removed files */
+ cgroup_populate_dir(cgrp, false, removed_mask);
drop_parsed_module_refcounts(opts.subsys_mask);
goto out_unlock;
}
- /* clear out any existing files and repopulate subsystem files */
- cgroup_clear_directory(cgrp->dentry, false, removed_mask);
/* re-populate subsystem files */
cgroup_populate_dir(cgrp, false, added_mask);
--
1.7.7.6
_______________________________________________
Containers mailing list
Containers@xxxxxxxxxxxxxxxxxxxxxxxxxx
https://lists.linuxfoundation.org/mailman/listinfo/containers
