From: Sage Weil <sage@newdream.net>
Date: Mon, 29 Mar 2010 04:22:50 +0000 (-0700)
Subject: ceph: fix use after free on mds __unregister_request
X-Git-Tag: v2.6.34-rc3~6^2~2
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=94aa8ae13db2ecf2ec1b4e65a65d3fe92b468e0e;p=profile%2Fcommon%2Fkernel-common.git

ceph: fix use after free on mds __unregister_request

There was a use after free in __unregister_request that would trigger
whenever the request map held the last reference.  This appears to have
triggered an oops during 'umount -f' when requests are being torn down.

Signed-off-by: Sage Weil <sage@newdream.net>
---

diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 5268d40..5c7920b 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -532,7 +532,6 @@ static void __unregister_request(struct ceph_mds_client *mdsc,
 	dout("__unregister_request %p tid %lld\n", req, req->r_tid);
 	rb_erase(&req->r_node, &mdsc->request_tree);
 	RB_CLEAR_NODE(&req->r_node);
-	ceph_mdsc_put_request(req);
 
 	if (req->r_unsafe_dir) {
 		struct ceph_inode_info *ci = ceph_inode(req->r_unsafe_dir);
@@ -541,6 +540,8 @@ static void __unregister_request(struct ceph_mds_client *mdsc,
 		list_del_init(&req->r_unsafe_dir_item);
 		spin_unlock(&ci->i_unsafe_lock);
 	}
+
+	ceph_mdsc_put_request(req);
 }
 
 /*