linux-stable.git/drivers/infiniband, branch v3.15.3

libiscsi, iser: Adjust data_length to include protection information

2014-07-01T03:14:00+00:00

commit d77e65350f2d82dfa0557707d505711f5a43c8fd upstream.

In case protection information exists over the wire
iscsi header data length is required to include it.
Use protection information aware scsi helpers to set
the correct transfer length.

In order to avoid breakage, remove iser transfer length
checks for each task as they are not always true and
somewhat redundant anyway.

Signed-off-by: Sagi Grimberg 
Reviewed-by: Mike Christie 
Acked-by: Mike Christie 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iscsi: Fix sendtargets response pdu for iser transport

2014-07-01T03:14:00+00:00

commit 22c7aaa57e80853b4904a46c18f97db0036a3b97 upstream.

In case the transport is iser we should not include the
iscsi target info in the sendtargets text response pdu.
This causes sendtargets response to include the target
info twice.

Modify iscsit_build_sendtargets_response to filter
transport types that don't match.

Signed-off-by: Sagi Grimberg 
Reported-by: Slava Shwartsman 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Fix a wrong dereference in case discovery session is over iser

2014-07-01T03:14:00+00:00

commit e0546fc1ba66c90cb38a5764357366267d3e58e4 upstream.

In case the discovery session is carried over iser, we can't
access the assumed network portal since the default portal is
used. In this case we don't really need to allocate the fastreg
pool, just prepare to the text pdu that will follow.

Signed-off-by: Sagi Grimberg 
Reported-by: Alex Tabachnik 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Gracefully reject T10-PI enabled connect request if not supported

2014-07-01T03:14:00+00:00

commit 5f80ff8eccba50832dcc640ac89add4c7fced963 upstream.

In case user chose to set T10-PI enable on the target while
the IB device does not support it, gracefully reject the request.

Reported-by: Slava Shwartsman 
Signed-off-by: Sagi Grimberg 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Wait for proper cleanup before unloading

2014-07-01T03:14:00+00:00

commit f5ebec9629cf78eeeea4b8258882a9f439ab2404 upstream.

disconnected_handler works are scheduled on system_wq.
When attempting to unload, first make sure all works
have cleaned up.

Signed-off-by: Sagi Grimberg 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Improve cm events handling

2014-07-01T03:14:00+00:00

commit 88c4015fda6d014392f76d3b1688347950d7a12d upstream.

There are 4 RDMA_CM events that all basically mean that
the user should teardown the IB connection:
- DISCONNECTED
- ADDR_CHANGE
- DEVICE_REMOVAL
- TIMEWAIT_EXIT

Only in DISCONNECTED/ADDR_CHANGE it makes sense to
call rdma_disconnect (send DREQ/DREP to our initiator).
So we keep the same teardown handler for all of them
but only indicate calling rdma_disconnect for the relevant
events.

This patch also removes redundant debug prints for each single
event.

v2 changes:
 - Call isert_disconnected_handler() for DEVICE_REMOVAL (Or + Sag)

Signed-off-by: Sagi Grimberg 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Fix hangs in connection teardown

2014-07-01T03:13:59+00:00

commit 9d49f5e284e700576f3b65f1e28dea8539da6661 upstream.

In ungraceful teardowns isert close flows seem racy such that
isert_wait_conn hangs as RDMA_CM_EVENT_DISCONNECTED never
gets invoked (no one called rdma_disconnect).

Both graceful and ungraceful teardowns will have rx flush errors
(isert posts a batch once connection is established). Once all
flush errors are consumed we invoke isert_wait_conn and it will
be responsible for calling rdma_disconnect. This way it can be
sure that rdma_disconnect was called and it won't wait forever.

This patch also removes the logout_posted indicator. either the
logout completion was consumed and no problem decrementing the
post_send_buf_count, or it was consumed as a flush error. no point
of keeping it for isert_wait_conn as there is no danger that
isert_conn will be accidentally removed while it is running.

(Drop unnecessary sleep_on_conn_wait_comp check in
 isert_cq_rx_comp_err - nab)

Signed-off-by: Sagi Grimberg 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Target/iser: Bail from accept_np if np_thread is trying to close

2014-07-01T03:13:59+00:00

commit e346ab343f4f58c12a96725c7b13df9cc2ad56f6 upstream.

In case np_thread state is in RESET/SHUTDOWN/EXIT states,
no point for isert to stall there as we may get a hang in
case no one will wake it up later.

Signed-off-by: Sagi Grimberg 
Signed-off-by: Nicholas Bellinger 
Signed-off-by: Greg Kroah-Hartman

Merge git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending

2014-06-07T22:01:39+00:00

Pull SCSI target fixes from Nicholas Bellinger:
 "Here are the remaining fixes for v3.15.

  This series includes:

   - iser-target fix for ImmediateData exception reference count bug
     (Sagi + nab)
   - iscsi-target fix for MC/S login + potential iser-target MRDSL
     buffer overrun (Santosh + Roland)
   - iser-target fix for v3.15-rc multi network portal shutdown
     regression (nab)
   - target fix for allowing READ_CAPCITY during ALUA Standby access
     state (Chris + nab)
   - target fix for NULL pointer dereference of alua_access_state for
     un-configured devices (Chris + nab)"

* git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending:
  target: Fix alua_access_state attribute OOPs for un-configured devices
  target: Allow READ_CAPACITY opcode in ALUA Standby access state
  iser-target: Fix multi network portal shutdown regression
  iscsi-target: Fix wrong buffer / buffer overrun in iscsi_change_param_value()
  iser-target: Add missing target_put_sess_cmd for ImmedateData failure

iser-target: Add missing target_put_sess_cmd for ImmedateData failure

2014-06-04T02:17:31+00:00

This patch addresses a bug where an early exception for SCSI WRITE
with ImmediateData=Yes was missing the target_put_sess_cmd() call
to drop the extra se_cmd->cmd_kref reference obtained during the
normal iscsit_setup_scsi_cmd() codepath execution.

This bug was manifesting itself during session shutdown within
isert_cq_rx_comp_err() where target_wait_for_sess_cmds() would
end up waiting indefinately for the last se_cmd->cmd_kref put to
occur for the failed SCSI WRITE + ImmediateData descriptors.

This fix follows what traditional iscsi-target code already does
for the same failure case within iscsit_get_immediate_data().

Reported-by: Sagi Grimberg 
Cc: Sagi Grimberg 
Cc: Or Gerlitz 
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger