linux-stable.git/drivers/scsi, branch v5.2.20

scsi: core: Reduce memory required for SCSI logging

2019-10-07T16:59:22+00:00

[ Upstream commit dccc96abfb21dc19d69e707c38c8ba439bba7160 ]

The data structure used for log messages is so large that it can cause a
boot failure. Since allocations from that data structure can fail anyway,
use kmalloc() / kfree() instead of that data structure.

See also https://bugzilla.kernel.org/show_bug.cgi?id=204119.
See also commit ded85c193a39 ("scsi: Implement per-cpu logging buffer") # v4.0.

Reported-by: Jan Palus 
Cc: Christoph Hellwig 
Cc: Hannes Reinecke 
Cc: Johannes Thumshirn 
Cc: Ming Lei 
Cc: Jan Palus 
Signed-off-by: Bart Van Assche 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin

scsi: implement .cleanup_rq callback

2019-10-05T11:14:03+00:00

[ Upstream commit b7e9e1fb7a9227be34ad4a5e778022c3164494cf ]

Implement .cleanup_rq() callback for freeing driver private part
of the request. Then we can avoid to leak this part if the request isn't
completed by SCSI, and freed by blk-mq or upper layer(such as dm-rq) finally.

Cc: Ewan D. Milne 
Cc: Bart Van Assche 
Cc: Hannes Reinecke 
Cc: Christoph Hellwig 
Cc: Mike Snitzer 
Cc: dm-devel@redhat.com
Cc: 
Fixes: 396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via blk_insert_cloned_request feedback")
Signed-off-by: Ming Lei 
Signed-off-by: Jens Axboe 
Signed-off-by: Sasha Levin

scsi: qla2xxx: Fix Relogin to prevent modifying scan_state flag

2019-10-05T11:14:01+00:00

commit 8b5292bcfcacf15182a77a973a98d310e76fd58b upstream.

Relogin fails to move forward due to scan_state flag indicating device is
not there. Before relogin process, Session delete process accidently
modified the scan_state flag.

[mkp: typos plus corrected Fixes: sha as reported by sfr]

Fixes: 2dee5521028c ("scsi: qla2xxx: Fix login state machine freeze")
Cc: stable@vger.kernel.org
Signed-off-by: Quinn Tran 
Signed-off-by: Himanshu Madhani 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Greg Kroah-Hartman

scsi: scsi_dh_rdac: zero cdb in send_mode_select()

2019-10-05T11:14:01+00:00

commit 57adf5d4cfd3198aa480e7c94a101fc8c4e6109d upstream.

cdb in send_mode_select() is not zeroed and is only partially filled in
rdac_failover_get(), which leads to some random data getting to the
device. Users have reported storage responding to such commands with
INVALID FIELD IN CDB. Code before commit 327825574132 was not affected, as
it called blk_rq_set_block_pc().

Fix this by zeroing out the cdb first.

Identified & fix proposed by HPE.

Fixes: 327825574132 ("scsi_dh_rdac: switch to scsi_execute_req_flags()")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20190904155205.1666-1-martin.wilck@suse.com
Signed-off-by: Martin Wilck 
Acked-by: Ales Novak 
Reviewed-by: Shane Seymour 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Greg Kroah-Hartman

scsi: lpfc: Mitigate high memory pre-allocation by SCSI-MQ

2019-09-10T09:35:20+00:00

[ Upstream commit 77ffd3465ba837e9dc714e17b014e77b2eae765a ]

When SCSI-MQ is enabled, the SCSI-MQ layers will do pre-allocation of MQ
resources based on shost values set by the driver. In newer cases of the
driver, which attempts to set nr_hw_queues to the cpu count, the
multipliers become excessive, with a single shost having SCSI-MQ
pre-allocation reaching into the multiple GBytes range.  NPIV, which
creates additional shosts, only multiply this overhead. On lower-memory
systems, this can exhaust system memory very quickly, resulting in a system
crash or failures in the driver or elsewhere due to low memory conditions.

After testing several scenarios, the situation can be mitigated by limiting
the value set in shost->nr_hw_queues to 4. Although the shost values were
changed, the driver still had per-cpu hardware queues of its own that
allowed parallelization per-cpu.  Testing revealed that even with the
smallish number for nr_hw_queues for SCSI-MQ, performance levels remained
near maximum with the within-driver affiinitization.

A module parameter was created to allow the value set for the nr_hw_queues
to be tunable.

Signed-off-by: Dick Kennedy 
Signed-off-by: James Smart 
Reviewed-by: Ming Lei 
Reviewed-by: Ewan D. Milne 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin

scsi: qla2xxx: Fix gnl.l memory leak on adapter init failure

2019-09-10T09:35:15+00:00

[ Upstream commit 26fa656e9a0cbccddf7db132ea020d2169dbe46e ]

If HBA initialization fails unexpectedly (exiting via probe_failed:), we
may fail to free vha->gnl.l. So that we don't attempt to double free, set
this pointer to NULL after a free and check for NULL at probe_failed: so we
know whether or not to call dma_free_coherent.

Signed-off-by: Bill Kuzeja 
Acked-by: Himanshu Madhani 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin

scsi: ufs: Fix NULL pointer dereference in ufshcd_config_vreg_hpm()

2019-08-29T06:30:22+00:00

commit 7c7cfdcf7f1777c7376fc9a239980de04b6b5ea1 upstream.

Fix the following BUG:

  [ 187.065689] BUG: kernel NULL pointer dereference, address: 000000000000001c
  [ 187.065790] RIP: 0010:ufshcd_vreg_set_hpm+0x3c/0x110 [ufshcd_core]
  [ 187.065938] Call Trace:
  [ 187.065959] ufshcd_resume+0x72/0x290 [ufshcd_core]
  [ 187.065980] ufshcd_system_resume+0x54/0x140 [ufshcd_core]
  [ 187.065993] ? pci_pm_restore+0xb0/0xb0
  [ 187.066005] ufshcd_pci_resume+0x15/0x20 [ufshcd_pci]
  [ 187.066017] pci_pm_thaw+0x4c/0x90
  [ 187.066030] dpm_run_callback+0x5b/0x150
  [ 187.066043] device_resume+0x11b/0x220

Voltage regulators are optional, so functions must check they exist
before dereferencing.

Note this issue is hidden if CONFIG_REGULATORS is not set, because the
offending code is optimised away.

Notes for stable:

The issue first appears in commit 57d104c153d3 ("ufs: add UFS power
management support") but is inadvertently fixed in commit 60f0187031c0
("scsi: ufs: disable vccq if it's not needed by UFS device") which in
turn was reverted by commit 730679817d83 ("Revert "scsi: ufs: disable vccq
if it's not needed by UFS device""). So fix applies v3.18 to v4.5 and
v5.1+

Fixes: 57d104c153d3 ("ufs: add UFS power management support")
Fixes: 730679817d83 ("Revert "scsi: ufs: disable vccq if it's not needed by UFS device"")
Cc: stable@vger.kernel.org
Signed-off-by: Adrian Hunter 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Greg Kroah-Hartman

scsi: qla2xxx: Fix possible fcport null-pointer dereferences

2019-08-25T14:10:30+00:00

[ Upstream commit e82f04ec6ba91065fd33a6201ffd7cab840e1475 ]

In qla2x00_alloc_fcport(), fcport is assigned to NULL in the error
handling code on line 4880:
    fcport = NULL;

Then fcport is used on lines 4883-4886:
    INIT_WORK(&fcport->del_work, qla24xx_delete_sess_fn);
	INIT_WORK(&fcport->reg_work, qla_register_fcport_fn);
	INIT_LIST_HEAD(&fcport->gnl_entry);
	INIT_LIST_HEAD(&fcport->list);

Thus, possible null-pointer dereferences may occur.

To fix these bugs, qla2x00_alloc_fcport() directly returns NULL
in the error handling code.

These bugs are found by a static analysis tool STCheck written by us.

Signed-off-by: Jia-Ju Bai 
Acked-by: Himanshu Madhani 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin

scsi: hpsa: correct scsi command status issue after reset

2019-08-25T14:10:30+00:00

[ Upstream commit eeebce1862970653cdf5c01e98bc669edd8f529a ]

Reviewed-by: Bader Ali - Saleh 
Reviewed-by: Scott Teel 
Reviewed-by: Scott Benesh 
Reviewed-by: Kevin Barnett 
Signed-off-by: Don Brace 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin

scsi: scsi_dh_alua: always use a 2 second delay before retrying RTPG

2019-08-16T08:11:06+00:00

[ Upstream commit 20122994e38aef0ae50555884d287adde6641c94 ]

Retrying immediately after we've received a 'transitioning' sense code is
pretty much pointless, we should always use a delay before retrying.  So
ensure the default delay is applied before retrying.

Signed-off-by: Hannes Reinecke 
Tested-by: Zhangguanghui 
Signed-off-by: Martin K. Petersen 
Signed-off-by: Sasha Levin