Commit 992f0b22 authored by Liang Zhen's avatar Liang Zhen Committed by Greg Kroah-Hartman

staging: lustre: take extra refcount in kiblnd_connreq_done

refcount taken by cmid is not reliable after kiblnd_connreq_done
released the glock because this connection is visible to other
threads, another thread can find and close this connection right
after kiblnd_connreq_done released the glock, if kiblnd_cm_callback
for RDMA_CM_EVENT_DISCONNECTED is called, it can release the
connection refcount taken by cmid. It means the connection could be
destroyed before kiblnd_connreq_done() finish operations on it.
Signed-off-by: default avatarLiang Zhen <liang.zhen@intel.com>
ntel-bug-id: https://jira.hpdd.intel.com/browse/LU-7210
Reviewed-on: http://review.whamcloud.com/17527Reviewed-by: default avatarDoug Oucharek <doug.s.oucharek@intel.com>
Reviewed-by: default avatarJames Simmons <uja.ornl@yahoo.com>
Tested-by: default avatarJames Simmons <uja.ornl@yahoo.com>
Reviewed-by: default avatarOleg Drokin <oleg.drokin@intel.com>
Signed-off-by: default avatarGreg Kroah-Hartman <gregkh@linuxfoundation.org>
parent a01fa108
...@@ -939,8 +939,6 @@ kiblnd_check_sends(kib_conn_t *conn) ...@@ -939,8 +939,6 @@ kiblnd_check_sends(kib_conn_t *conn)
kiblnd_queue_tx_locked(tx, conn); kiblnd_queue_tx_locked(tx, conn);
} }
kiblnd_conn_addref(conn); /* 1 ref for me.... (see b21911) */
for (;;) { for (;;) {
int credit; int credit;
...@@ -966,8 +964,6 @@ kiblnd_check_sends(kib_conn_t *conn) ...@@ -966,8 +964,6 @@ kiblnd_check_sends(kib_conn_t *conn)
} }
spin_unlock(&conn->ibc_lock); spin_unlock(&conn->ibc_lock);
kiblnd_conn_decref(conn); /* ...until here */
} }
static void static void
...@@ -2131,6 +2127,16 @@ kiblnd_connreq_done(kib_conn_t *conn, int status) ...@@ -2131,6 +2127,16 @@ kiblnd_connreq_done(kib_conn_t *conn, int status)
return; return;
} }
/**
* refcount taken by cmid is not reliable after I released the glock
* because this connection is visible to other threads now, another
* thread can find and close this connection right after I released
* the glock, if kiblnd_cm_callback for RDMA_CM_EVENT_DISCONNECTED is
* called, it can release the connection refcount taken by cmid.
* It means the connection could be destroyed before I finish my
* operations on it.
*/
kiblnd_conn_addref(conn);
write_unlock_irqrestore(&kiblnd_data.kib_global_lock, flags); write_unlock_irqrestore(&kiblnd_data.kib_global_lock, flags);
/* Schedule blocked txs */ /* Schedule blocked txs */
...@@ -2146,6 +2152,8 @@ kiblnd_connreq_done(kib_conn_t *conn, int status) ...@@ -2146,6 +2152,8 @@ kiblnd_connreq_done(kib_conn_t *conn, int status)
/* schedule blocked rxs */ /* schedule blocked rxs */
kiblnd_handle_early_rxs(conn); kiblnd_handle_early_rxs(conn);
kiblnd_conn_decref(conn);
} }
static void static void
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment