Vaughan Cao [Thu, 29 Aug 2013 02:00:38 +0000 (10:00 +0800)]
[SCSI] sg: checking sdp->detached isn't protected when open
@detached is set under the protection of sg_index_lock. Without getting the
lock, new sfp will be added during sg removal and there is no chance for it
to be picked out. So check with sg_index_lock held in sg_add_sfp().
Signed-off-by: Vaughan Cao <vaughan.cao@oracle.com> Acked-by: Douglas Gilbert <dgilbert@interlog.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Vaughan Cao [Thu, 29 Aug 2013 02:00:37 +0000 (10:00 +0800)]
[SCSI] sg: no need sg_open_exclusive_lock
Open exclusive check is protected by o_sem, no need sg_open_exclusive_lock.
@exclude is used to record which type of rwsem we are holding.
Signed-off-by: Vaughan Cao <vaughan.cao@oracle.com> Acked-by: Douglas Gilbert <dgilbert@interlog.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Vaughan Cao [Thu, 29 Aug 2013 02:00:36 +0000 (10:00 +0800)]
[SCSI] sg: use rwsem to solve race during exclusive open
A race condition may happen if two threads are both trying to open the same sg
with O_EXCL simultaneously. It's possible that they both find fsds list is
empty and get_exclude(sdp) returns 0, then they both call set_exclude() and
break out from wait_event_interruptible and resume open.
Now use rwsem to protect this process. Exclusive open gets write lock and
others get read lock. The lock will be held until file descriptor is closed.
This also leads 'exclude' only a status rather than a check mark.
Signed-off-by: Vaughan Cao <vaughan.cao@oracle.com> Acked-by: Douglas Gilbert <dgilbert@interlog.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Akinobu Mita [Mon, 26 Aug 2013 13:08:41 +0000 (22:08 +0900)]
[SCSI] scsi_debug: fix logical block provisioning support when unmap_alignment != 0
Commit b90ebc3d5c41c9164ae04efd2e4f8204c2a186f1 ("[SCSI] scsi_debug:
fix logical block provisioning support") fixed several issues with
logical block provisioning support, but it still doesn't properly fix
the cases when unmap_alignment > 0.
For example, load scsi_debug module with the following module parameters
and make all blocks mapped by filling the storage with zero.
The problem is in map_index_to_lba(), which should return the first
LBA which is corresponding to a given index of provisioning map
(map_storep).
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Acked-by: "Martin K. Petersen" <martin.petersen@oracle.com> Acked-by: Douglas Gilbert <dgilbert@interlog.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Akinobu Mita [Mon, 26 Aug 2013 13:08:40 +0000 (22:08 +0900)]
[SCSI] scsi_debug: fix endianness bug in sdebug_build_parts()
With module parameter num_parts > 0, partition table is built on the
ramdisk storage when loading the driver. Unfortunately, there is an
endianness bug in sdebug_build_parts(). So the partition table is not
correctly initialized on big-endian systems.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com> Tested-by: Martin Peschke <mpeschke@linux.vnet.ibm.com> Acked-by: Douglas Gilbert <dgilbert@interlog.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Sawan Chandak [Tue, 27 Aug 2013 05:37:51 +0000 (01:37 -0400)]
[SCSI] qla2xxx: Add a new interface to update versions.
On any Adapter when we flash through FC/FCoE without card reset option it still
shows the Running FW version in Flashed FW version. This new interface will be
used by API to instruct the driver to update its cache versions.
Chad Dupuis [Tue, 27 Aug 2013 05:37:50 +0000 (01:37 -0400)]
[SCSI] qla2xxx: Move queue depth ramp down message to i/o debug level.
Unless there is a need to observe them, the queue depth ramp up/down messages
are a nuisance and may cause the system to become unresponsive so move it a
non-default logging level.
Signed-off-by: Chad Dupuis <chad.dupuis@qlogic.com> Signed-off-by: Saurav Kashyap <saurav.kashyap@qlogic.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Earlier IDC-TIME-EXTEND aen was a nop and ignored by driver.
We now have to handle the aen so that other protocol drivers can use time
extension during some loopback operations.
[SCSI] qla2xxx: Print some variables to hexadecimal string via %*phN format
The patch changes a specifier used to output some variables. Instead of using
stack for each byte the '%*ph[CN]' allows to take a one pointer and prints
entire buffer as a hexadecimal string with the separator ':' or ''.
Steffen Maier [Thu, 22 Aug 2013 15:49:34 +0000 (17:49 +0200)]
[SCSI] zfcp: enable FCP hardware data router by default
Enabling the data router support by default
can increase performance in certain situations.
It is safe to do so and tolerated in LPAR and under z/VM
in case there is no data router support in that environment.
Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com> Reviewed-by: Martin Peschke <mpeschke@linux.vnet.ibm.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Martin Peschke [Thu, 22 Aug 2013 15:49:33 +0000 (17:49 +0200)]
[SCSI] zfcp: dead code removal
Get rid of unused function zfcp_fsf_get_req and corresponding
prototype definition.
Commit a54ca0f62f953898b05549391ac2a8a4dad6482b in v2.6.28
"[SCSI] zfcp: Redesign of the debug tracing for HBA records."
accidentally introduced this code which was dead in the first place.
Signed-off-by: Martin Peschke <mpeschke@linux.vnet.ibm.com> Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Sreekanth Reddy [Wed, 14 Aug 2013 12:53:20 +0000 (18:23 +0530)]
[SCSI] mpt3sas: Added a driver module parameter max_msix_vectors
Added a driver module parameter max_msix_vectors. Using this module parameter
the maximum number of MSI-X vectors could be set.
The number of MSI-X vectors used would be the minimum of MSI-X vectors
supported by the HBA, the number of CPU cores and the value set to
max_msix_vectors module parameter.
The default value of this module parameter is set to 8. The default value of
this parameter is set to 8 inorder to reduce the amount of memory required for
Reply Descriptor Post queue. This is because with the higher MSI-X vectors,
some times kernel is not able to allocate the requested amount of memory and
crash is observed. To overcome this problem, the default value is set to 8.
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Nilesh Javali [Fri, 16 Aug 2013 11:03:03 +0000 (07:03 -0400)]
[SCSI] qla4xxx: Implementation of ACB configuration during Loopback for ISP8042
While loopback diagnostic is in progress, disable the ACB which resets
all the active connections to target. Disable ACB would filter out all
the DHCP multicast and broadcast packets which otherwise cause the
diagnostic test to take longer time to complete or failures in some
other cases.
[SCSI] mpt2sas: Fix for kernel panic when driver loads with HBA connected to non LUN 0 configured expander
With some enclosures when LUN 0 is not created but LUN 1 or LUN X is created
then SCSI scan procedure calls target_alloc, slave_alloc call back functions
for LUN 0 and slave_destory() for same LUN 0.
In these kind of cases within slave_destroy, pointer to scsi_target in
_sas_device structure is set to NULL, following which when slave_alloc for
LUN 1 is called then starget would not be set properly for this LUN.
So, scsi_target pointer pointing to NULL value would lead to a crash later
in the discovery procedure.
To solve this issue set the sas_device's scsi_target pointer to scsi_device's
scsi_target if it is NULL earlier in slave_alloc callback function.
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] mpt2sas: when Async scanning is enabled then while scanning, devices are removed but their transport layer entries are not removed
When Async scanning mode is enabled and device scanning is in progress then
devices should not be removed. But in actuality, devices are removed but
their transport layer entries are not removed. This causes error to add
the same device to the transport layer after host reset or diagnostic
reset.
So, in this patch, modified the code in such a way that device is not removed
when Async scanning mode is enabled and device scanning is in progress.
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] mpt2sas: Infinite loop can occur if MPI2_IOCSTATUS_CONFIG_INVALID_PAGE is not returned
Infinite loop can occur if IOCStatus is not equal to
MPI2_IOCSTATUS_CONFIG_INVALID_PAGE value in the while loops in functions
_scsih_search_responding_sas_devices,
_scsih_search_responding_raid_devices and
_scsih_search_responding_expanders
So, Instead of checking for MPI2_IOCSTATUS_CONFIG_INVALID_PAGE value,
in this patch code is modified to check for IOCStatus not equals to
MPI2_IOCSTATUS_SUCCESS to break the while loop.
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] mpt2sas: Change in MPI2_RAID_ACTION_SYSTEM_SHUTDOWN_INITIATED notification methodology
The intent of this patch is to perform a graceful shutdown of target drives even if
volume doesn't exits. Changes done in this patch
1. Removed the check for the presence of volumes before sending down
MPI2_RAID_ACTION_SYSTEM_SHUTDOWN_INITIATED. Therefore, this RAID action
would be sent if the card is IR Firmware.
2. The MPI2_RAID_ACTION_SYSTEM_SHUTDOWN_INITIATED is sent even when the
system undergoes suspend (in addition to remove/shutdown which was already
present)
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] mpt2sas: Null pointer deference possibility in mpt2sas_ctl_event_callback function
Added a check to identify if mpi_reply is NULL in mpt2sas_ctl_event_callback()
and return without proceeding if it is the case.
Also modified the following functions to return void instead of 0 or 1
as returning those values from events perspective doesn't make sense.
* _base_async_event()
* mpt2sas_ctl_event_callback()
* mpt2sas_scsih_event_callback()
Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
This is caused by
enqueue_cmd_and_start_io()->
set_performant_mode()->
smp_processor_id()
Which if you have debugging enabled calls debug_processor_id() and triggers the warning.
The code here is
c->Header.ReplyQueue = smp_processor_id() % h->nreply_queues;
Since it is not critical that the code complete on the same processor,
but the cpu is a hint used in generating the ReplyQueue and will still work if
the cpu migrates or is preempted, it is safe to use the raw_smp_processor_id()
to surpress the false positve warning.
Signed-off-by: John Kacur <jkacur@redhat.com> Acked-by: Stephen Cameron <scameron@beardog.cce.hp.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Or Gerlitz [Thu, 8 Aug 2013 10:44:29 +0000 (13:44 +0300)]
[SCSI] IB/iser: Add Discovery support
To run discovery over iSER we need to advertize the CAP_TEXT_NEGO capability
towards user space. Also need to make sure the login RX buffer is posted when
SendTargets TEXT PDUs are sent. For that end, we use a setting of the
ISCSI_PARAM_DISCOVERY_SESS iscsi param as an indication that this is
discovery session.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Reviewed-by: Mike Christie <michaelc@cs.wisc.edu> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Yijing Wang [Thu, 8 Aug 2013 13:10:21 +0000 (21:10 +0800)]
[SCSI] pm8001: clean up unnecessary MSI/MSI-X capability find
PCI core will initialize device MSI/MSI-X capability in
pci_msi_init_pci_dev(). So device driver should use
pci_dev->msi_cap/msix_cap to determine whether the device
support MSI/MSI-X instead of using
pci_find_capability(pci_dev, PCI_CAP_ID_MSI/MSIX).
Access to PCIe device config space again will consume more time.
Signed-off-by: Yijing Wang <wangyijing@huawei.com> Acked-by: lindar_liu <lindar_liu@usish.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Joe Lawrence [Thu, 8 Aug 2013 20:45:39 +0000 (16:45 -0400)]
[SCSI] mpt3sas: fix cleanup on controller resource mapping failure
If mpt3sas_base_map_resources takes an early error path then its
counterpart, mpt3sas_base_free_resources needs to be careful about
cleaning up:
1 - _base_mask_interrupts and _base_make_ioc_ready require memory
mapped I/O registers, make sure that this is true.
2 - _base_free_irq iterates over the adapter's reply_queue_list, so
move this list head initialization out of _base_enable_msix to
_scsih_probe so this will always be safe.
3 - check that the controller PCI device and its BARs have been
enabled before disabling them.
Signed-off-by: Joe Lawrence <joe.lawrence@stratus.com> Acked-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Joe Lawrence [Thu, 8 Aug 2013 20:45:38 +0000 (16:45 -0400)]
[SCSI] mpt2sas: fix cleanup on controller resource mapping failure
If mpt2sas_base_map_resources takes an early error path then its
counterpart, mpt2sas_base_free_resources needs to be careful about
cleaning up:
1 - _base_mask_interrupts and _base_make_ioc_ready require memory
mapped I/O registers, make sure that this is true.
2 - _base_free_irq iterates over the adapter's reply_queue_list, so
move this list head initialization out of _base_enable_msix to
_scsih_probe so this will always be safe.
3 - check that the controller PCI device and its BARs have been
enabled before disabling them.
Signed-off-by: Joe Lawrence <joe.lawrence@stratus.com> Acked-by: Sreekanth Reddy <Sreekanth.Reddy@lsi.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Ewan D. Milne [Thu, 8 Aug 2013 19:07:48 +0000 (15:07 -0400)]
[SCSI] Generate uevents on certain unit attention codes
Generate a uevent when the following Unit Attention ASC/ASCQ
codes are received:
2A/01 MODE PARAMETERS CHANGED
2A/09 CAPACITY DATA HAS CHANGED
38/07 THIN PROVISIONING SOFT THRESHOLD REACHED
3F/03 INQUIRY DATA HAS CHANGED
3F/0E REPORTED LUNS DATA HAS CHANGED
Log kernel messages when the following Unit Attention ASC/ASCQ
codes are received that are not as specific as those above:
2A/xx PARAMETERS CHANGED
3F/xx TARGET OPERATING CONDITIONS HAVE CHANGED
Added logic to set expecting_lun_change for other LUNs on the target
after REPORTED LUNS DATA HAS CHANGED is received, so that duplicate
uevents are not generated, and clear expecting_lun_change when a
REPORT LUNS command completes, in accordance with the SPC-3
specification regarding reporting of the 3F 0E ASC/ASCQ UA.
[jejb: remove SPC3 test in scsi_report_lun_change and some docbook fixes and
unused variable fix, both reported by Fengguang Wu] Signed-off-by: Ewan D. Milne <emilne@redhat.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Tomas Henzl [Thu, 1 Aug 2013 13:14:52 +0000 (15:14 +0200)]
[SCSI] hpsa: remove unneeded variable
Signed-off-by: Tomas Henzl <thenzl@redhat.com> Acked-by: Stephen M. Cameron <scameron@beardog.cce.hp.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Tomas Henzl [Thu, 1 Aug 2013 13:14:00 +0000 (15:14 +0200)]
[SCSI] hpsa: fix a race in cmd_free/scsi_done
When the driver calls scsi_done and after that frees it's internal
preallocated memory it can happen that a new job is enqueud before
the memory is freed. The allocation fails and the message
"cmd_alloc returned NULL" is shown.
Patch below fixes it by moving cmd->scsi_done after cmd_free.
Signed-off-by: Tomas Henzl <thenzl@redhat.com> Acked-by: Stephen M. Cameron <scameron@beardog.cce.hp.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Wolfram Sang [Mon, 29 Jul 2013 19:06:05 +0000 (00:36 +0530)]
[SCSI] ufs: don't check resource with devm_ioremap_resource
devm_ioremap_resource does sanity checks on the given resource. No need to
duplicate this in the driver.
Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Akinobu Mita [Mon, 29 Jul 2013 19:06:03 +0000 (00:36 +0530)]
[SCSI] ufs: don't stop controller before scsi_remove_host()
scsi_remove_host() sends SYNCHRONIZE CACHE commands for write cache
enabled scsi disk devices. So stopping controller working shouldn't
be done before scsi_remove_host().
Signed-off-by: Akinobu Mita <mita@fixstars.com> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Akinobu Mita [Mon, 29 Jul 2013 19:06:02 +0000 (00:36 +0530)]
[SCSI] ufs: don't disable_irq() if the IRQ can be shared among devices
When removing the UFS driver, disable_irq() is called and the IRQ is
not enabled again. Unfortunately, the IRQ is requested with IRQF_SHARED
and it can be shared among several devices. So disabling the IRQ in
this way is just breaking other devices which are sharing the IRQ.
Signed-off-by: Akinobu Mita <mita@fixstars.com> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
Akinobu Mita [Mon, 29 Jul 2013 19:06:01 +0000 (00:36 +0530)]
[SCSI] ufshcd-pci: release ioremapped region during removing driver
Before commit 2953f850c3b80bdca004967c83733365d8aa0aa2 ("[SCSI] ufs:
use devres functions for ufshcd"), UFSHCI register was ioremapped by
each glue-driver (ufshcd-pltfrm and ufshcd-pci) during probing and it
was iounmapped by core-driver during removing driver. The commit
converted ufshcd-pltfrm to use devres functions, but it didn't convert
ufshcd-pci.
Therefore, the change causes ufshcd-pci driver not to iounmap UFSHCI
register region during removing driver. This fixes it by converting
ufshcd-pci to use devres functions.
Signed-off-by: Akinobu Mita <mita@fixstars.com> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] ufs: Add runtime PM support for UFS host controller driver
Add runtime PM helpers to suspend/resume UFS controller at runtime.
Enable runtime PM by default for pci and platform drivers as the
initialized hardware can suspend if it is not used after bootup.
Signed-off-by: Sujit Reddy Thumma <sthumma@codeaurora.org> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] ufs: Add support for host assisted background operations
Background operations in the UFS device can be disabled by
the host to reduce the response latency of transfer requests.
Add support for enabling/disabling the background operations
during runtime suspend/resume of the device.
If the device is in critical need of BKOPS it will raise an
URGENT_BKOPS exception which should be handled by the host to
make sure the device performs as expected.
During bootup, the BKOPS is enabled in the device by default.
The disable of BKOPS is supported only when the driver supports
runtime suspend/resume operations as the runtime PM framework
provides a way to determine the device idleness and hence BKOPS
can be managed effectively. During runtime resume the BKOPS is
disabled to reduce latency and during runtime suspend the BKOPS
is enabled to allow device to carry out idle time BKOPS.
In some cases where the BKOPS is disabled during runtime resume
and due to continuous data transfers the runtime suspend is not
triggered, the BKOPS is enabled when the device raises a level-2
exception (outstanding operations - performance impact).
Signed-off-by: Sujit Reddy Thumma <sthumma@codeaurora.org> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
[SCSI] ufs: Set fDeviceInit flag to initiate device initialization
Allow UFS device to complete its initialization and accept
SCSI commands by setting fDeviceInit flag. The device may take
time for this operation and hence the host should poll until
fDeviceInit flag is toggled to zero. This step is mandated by
UFS device specification for device initialization completion.
Signed-off-by: Dolev Raviv <draviv@codeaurora.org> Signed-off-by: Sujit Reddy Thumma <sthumma@codeaurora.org> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
As part of device initialization sequence, sending NOP OUT UPIU and
waiting for NOP IN UPIU response is mandatory. This confirms that the
device UFS Transport (UTP) layer is functional and the host can configure
the device with further commands. Add support for sending NOP OUT UPIU to
check the device connection path and test whether the UTP layer on the
device side is functional during initialization.
A tag is acquired from the SCSI tag map space in order to send the device
management command. When the tag is acquired by internal command the scsi
command is rejected with host busy flag in order to requeue the request.
To avoid frequent collisions between internal commands and scsi commands
the device management command tag is allocated in the opposite direction
w.r.t block layer tag allocation.
Signed-off-by: Sujit Reddy Thumma <sthumma@codeaurora.org> Signed-off-by: Dolev Raviv <draviv@codeaurora.org> Signed-off-by: Santosh Y <santoshsy@gmail.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
projects / linux-beck.git / log