[SCSI] Revert "[SCSI] fix async probe regression"
authorDan Williams <dan.j.williams@intel.com>
Tue, 10 Jul 2012 02:33:45 +0000 (19:33 -0700)
committerJames Bottomley <JBottomley@Parallels.com>
Fri, 20 Jul 2012 08:25:56 +0000 (09:25 +0100)
This reverts commit 43a8d39d0137612c336aa8bbb2cb886a79772ffb.

Commit 43a8d39d fixed the fact that wait_for_device_probe() was unable
to flush sd probe work.  Now that sd probe work is once again flushable
via wait_for_device_probe() this workaround is no longer needed.

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Tested-by: Eldad Zack <eldad@fogrefinery.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
drivers/scsi/scsi_scan.c

index a0bc6637265469eb2ffeea7fe9287d28d96f8562..56a93794c470ae99d603426f15eaccb1dd3ae727 100644 (file)
@@ -147,7 +147,7 @@ int scsi_complete_async_scans(void)
 
        do {
                if (list_empty(&scanning_hosts))
-                       goto out;
+                       return 0;
                /* If we can't get memory immediately, that's OK.  Just
                 * sleep a little.  Even if we never get memory, the async
                 * scans will finish eventually.
@@ -179,11 +179,8 @@ int scsi_complete_async_scans(void)
        }
  done:
        spin_unlock(&async_scan_lock);
-       kfree(data);
-
- out:
-       async_synchronize_full_domain(&scsi_sd_probe_domain);
 
+       kfree(data);
        return 0;
 }