summaryrefslogtreecommitdiffstats
path: root/fs/namespace.c
diff options
context:
space:
mode:
authorMiklos Szeredi <mszeredi@suse.cz>2010-10-05 12:31:09 +0200
committerAl Viro <viro@zeniv.linux.org.uk>2010-10-25 21:24:16 -0400
commitbe1a16a0ae29a7c90081a657b64aa51cb1a65a27 (patch)
tree18bef9178b1142d839b906ce3bf38c92eb650ee8 /fs/namespace.c
parent89b0fc38cca4e6c92a90b58960881ffc5dddd89c (diff)
downloadop-kernel-dev-be1a16a0ae29a7c90081a657b64aa51cb1a65a27.zip
op-kernel-dev-be1a16a0ae29a7c90081a657b64aa51cb1a65a27.tar.gz
vfs: fix infinite loop caused by clone_mnt race
If clone_mnt() happens while mnt_make_readonly() is running, the cloned mount might have MNT_WRITE_HOLD flag set, which results in mnt_want_write() spinning forever on this mount. Needs CAP_SYS_ADMIN to trigger deliberately and unlikely to happen accidentally. But if it does happen it can hang the machine. Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Diffstat (limited to 'fs/namespace.c')
-rw-r--r--fs/namespace.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/fs/namespace.c b/fs/namespace.c
index 7ca5182..8a415c9 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -595,7 +595,7 @@ static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root,
goto out_free;
}
- mnt->mnt_flags = old->mnt_flags;
+ mnt->mnt_flags = old->mnt_flags & ~MNT_WRITE_HOLD;
atomic_inc(&sb->s_active);
mnt->mnt_sb = sb;
mnt->mnt_root = dget(root);
OpenPOWER on IntegriCloud