[SCSI] libfc: Fix incorrect locking and unlocking in FCP
authorRobert Love <robert.w.love@intel.com>
Wed, 1 Dec 2010 00:18:23 +0000 (16:18 -0800)
committerJames Bottomley <James.Bottomley@suse.de>
Tue, 21 Dec 2010 18:24:20 +0000 (12:24 -0600)
The error handler grabs the si->scsi_queue_lock, but
in the case where the fsp pointer is NULL it releases
the scsi_host lock. This can lead to a variety of
system hangs depending on which is used first- the
scsi_host lock or the scsi_queue_lock.

This patch simply unlocks the correct lock when fcp
is NULL.

Signed-off-by: Robert Love <robert.w.love@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
drivers/scsi/libfc/fc_fcp.c

index 2924363d142bc4affc6422de1272e73a70957486..b9ad74de76da61e9392039fd00785b4082a6fca5 100644 (file)
@@ -2004,7 +2004,7 @@ int fc_eh_abort(struct scsi_cmnd *sc_cmd)
        fsp = CMD_SP(sc_cmd);
        if (!fsp) {
                /* command completed while scsi eh was setting up */
-               spin_unlock_irqrestore(lport->host->host_lock, flags);
+               spin_unlock_irqrestore(&si->scsi_queue_lock, flags);
                return SUCCESS;
        }
        /* grab a ref so the fsp and sc_cmd cannot be relased from under us */