tcm_qla2xxx: Fix residual for underrun commands that fail
authorRoland Dreier <roland@purestorage.com>
Wed, 5 Jun 2013 16:54:17 +0000 (09:54 -0700)
committerNicholas Bellinger <nab@linux-iscsi.org>
Fri, 14 Jun 2013 08:12:56 +0000 (01:12 -0700)
Suppose an initiator sends a DATA IN command with an allocation length
shorter than the FC transfer length -- we get a target message like

    TARGET_CORE[qla2xxx]: Expected Transfer Length: 256 does not match SCSI CDB Length: 0 for SAM Opcode: 0x12

In that case, the target core adjusts the data_length and sets
se_cmd->residual_count for the underrun.  But now suppose that command
fails and we end up in tcm_qla2xxx_queue_status() -- that function
unconditionally overwrites residual_count with the already adjusted
data_length, and the initiator will burp with a message like

    qla2xxx [0000:00:06.0]-301d:0: Dropped frame(s) detected (0x100 of 0x100 bytes).

Fix this by adding on to the existing underflow residual count instead.

Signed-off-by: Roland Dreier <roland@purestorage.com>
Cc: Giridhar Malavali <giridhar.malavali@qlogic.com>
Cc: Chad Dupuis <chad.dupuis@qlogic.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
drivers/scsi/qla2xxx/tcm_qla2xxx.c

index 7a3870f..66b0b26 100644 (file)
@@ -688,8 +688,12 @@ static int tcm_qla2xxx_queue_status(struct se_cmd *se_cmd)
                 * For FCP_READ with CHECK_CONDITION status, clear cmd->bufflen
                 * for qla_tgt_xmit_response LLD code
                 */
+               if (se_cmd->se_cmd_flags & SCF_OVERFLOW_BIT) {
+                       se_cmd->se_cmd_flags &= ~SCF_OVERFLOW_BIT;
+                       se_cmd->residual_count = 0;
+               }
                se_cmd->se_cmd_flags |= SCF_UNDERFLOW_BIT;
-               se_cmd->residual_count = se_cmd->data_length;
+               se_cmd->residual_count += se_cmd->data_length;
 
                cmd->bufflen = 0;
        }