scsi: storvsc: Fix scsi_cmd error assignments in storvsc_handle_error
authorCathy Avery <cavery@redhat.com>
Tue, 19 Dec 2017 18:32:48 +0000 (13:32 -0500)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 3 Mar 2018 09:24:28 +0000 (10:24 +0100)
commitb7b0385937fa163b8d8e68f38d90938143ef4feb
tree7ab1351222a25799034d5ece1220f8f12f39d402
parente2f52fa3f8d6fa949afebbb4dc24f71bc0785803
scsi: storvsc: Fix scsi_cmd error assignments in storvsc_handle_error

[ Upstream commit d1b8b2391c24751e44f618fcf86fb55d9a9247fd ]

When an I/O is returned with an srb_status of SRB_STATUS_INVALID_LUN
which has zero good_bytes it must be assigned an error. Otherwise the
I/O will be continuously requeued and will cause a deadlock in the case
where disks are being hot added and removed. sd_probe_async will wait
forever for its I/O to complete while holding scsi_sd_probe_domain.

Also returning the default error of DID_TARGET_FAILURE causes multipath
to not retry the I/O resulting in applications receiving I/O errors
before a failover can occur.

Signed-off-by: Cathy Avery <cavery@redhat.com>
Signed-off-by: Long Li <longli@microsoft.com>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <alexander.levin@verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/scsi/storvsc_drv.c