Ocfs2 devel - Dec 2014 - [PATCH 1/2] ocfs2: fix uninitialized variable access

Variable "why" is not yet initialized at line 615, fix it.

Signed-off-by: Junxiao Bi <junxiao.bi at oracle.com>
---
 fs/ocfs2/file.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 3950693..245db4f 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -569,7 +569,7 @@ static int __ocfs2_extend_allocation(struct inode *inode,
u32 logical_start,
 	handle_t *handle = NULL;
 	struct ocfs2_alloc_context *data_ac = NULL;
 	struct ocfs2_alloc_context *meta_ac = NULL;
-	enum ocfs2_alloc_restarted why;
+	enum ocfs2_alloc_restarted why = RESTART_NONE;
 	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 	struct ocfs2_extent_tree et;
 	int did_quota = 0;
-- 
1.7.9.5

O2NET_CONN_IDLE_DELAY is not defined, connection attempts will not be
canceled due to timeout.

Signed-off-by: Junxiao Bi <junxiao.bi at oracle.com>
---
 fs/ocfs2/cluster/tcp_internal.h |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/fs/ocfs2/cluster/tcp_internal.h b/fs/ocfs2/cluster/tcp_internal.h
index dc02436..b95e7df 100644
--- a/fs/ocfs2/cluster/tcp_internal.h
+++ b/fs/ocfs2/cluster/tcp_internal.h
@@ -107,12 +107,12 @@ struct o2net_node {
 	struct list_head		nn_status_list;
 
 	/* connects are attempted from when heartbeat comes up until either hb
-	 * goes down, the node is unconfigured, no connect attempts succeed
-	 * before O2NET_CONN_IDLE_DELAY, or a connect succeeds.  connect_work
-	 * is queued from set_nn_state both from hb up and from itself if a
-	 * connect attempt fails and so can be self-arming.  shutdown is
-	 * careful to first mark the nn such that no connects will be attempted
-	 * before canceling delayed connect work and flushing the queue. */
+	 * goes down, the node is unconfigured, or a connect succeeds.
+	 * connect_work is queued from set_nn_state both from hb up and from
+	 * itself if a connect attempt fails and so can be self-arming.
+	 * shutdown is careful to first mark the nn such that no connects will
+	 * be attempted before canceling delayed connect work and flushing the
+	 * queue. */
 	struct delayed_work		nn_connect_work;
 	unsigned long			nn_last_connect_attempt;
 
-- 
1.7.9.5

On Thu, 25 Dec 2014 13:52:16 +0800 Junxiao Bi <junxiao.bi at oracle.com>
wrote:
> Variable "why" is not yet initialized at line 615, fix it.
> 
> ...
>
> --- a/fs/ocfs2/file.c
> +++ b/fs/ocfs2/file.c
> @@ -569,7 +569,7 @@ static int __ocfs2_extend_allocation(struct inode
*inode, u32 logical_start,
>  	handle_t *handle = NULL;
>  	struct ocfs2_alloc_context *data_ac = NULL;
>  	struct ocfs2_alloc_context *meta_ac = NULL;
> -	enum ocfs2_alloc_restarted why;
> +	enum ocfs2_alloc_restarted why = RESTART_NONE;
>  	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
>  	struct ocfs2_extent_tree et;
>  	int did_quota = 0;
Oh geeze, are you really sure about this?  __ocfs2_extend_allocation()
is as clear as mud.  What happens when ocfs2_add_inode_data() returns
-EAGAIN and leaves *reason_ret unwritten to?

What are the runtime effects of this bug?