[SCSI] zfcp: revert previous patch for sbal counting

The current sbal counting can be wrong if a fsf request is
waiting for free sbals and at the same time qdio request queue
is shutdown and re-opened. Revering a previous patch fixes this
issue.

Signed-off-by: Martin Petermann <martin.petermann@de.ibm.com>
Signed-off-by: Christof Schmitt <christof.schmitt@de.ibm.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>

diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c
index bd2b23766894505991ce3bfda967dbd18a84a79d..3f365ace7711cafcb6c3cd341076d1bbb3a8e737 100644 (file)
--- a/drivers/s390/scsi/zfcp_fsf.c
+++ b/drivers/s390/scsi/zfcp_fsf.c
@@ -645,30 +645,30 @@ static void zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *req)
        }
 }
 
-static int zfcp_fsf_req_sbal_get(struct zfcp_adapter *adapter)
-       __releases(&adapter->req_q_lock)
-       __acquires(&adapter->req_q_lock)
+static int zfcp_fsf_sbal_check(struct zfcp_adapter *adapter)
 {
        struct zfcp_qdio_queue *req_q = &adapter->req_q;
-       long ret;
 
-       if (atomic_read(&req_q->count) <= -REQUEST_LIST_SIZE)
-               return -EIO;
-       if (atomic_read(&req_q->count) > 0)
-               return 0;
+       spin_lock_bh(&adapter->req_q_lock);
+       if (atomic_read(&req_q->count))
+               return 1;
+       spin_unlock_bh(&adapter->req_q_lock);
+       return 0;
+}
+
+static int zfcp_fsf_req_sbal_get(struct zfcp_adapter *adapter)
+{
+       long ret;
 
-       atomic_dec(&req_q->count);
        spin_unlock_bh(&adapter->req_q_lock);
        ret = wait_event_interruptible_timeout(adapter->request_wq,
-                                       atomic_read(&req_q->count) >= 0,
-                                       5 * HZ);
-       spin_lock_bh(&adapter->req_q_lock);
-       atomic_inc(&req_q->count);
-
+                              zfcp_fsf_sbal_check(adapter), 5 * HZ);
        if (ret > 0)
                return 0;
        if (!ret)
                atomic_inc(&adapter->qdio_outb_full);
+
+       spin_lock_bh(&adapter->req_q_lock);
        return -EIO;
 }