Function pci_enable_msix() should not be called in case
it threw a negative errno from a previous call.
Signed-off-by: NAlexander Gordeev <agordeev@redhat.com>
Acked-by: NAnil Gurumurthy <anil.gurumurthy@qlogic.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

As result of deprecation of MSI-X/MSI enablement functions
pci_enable_msix() and pci_enable_msi_block() all drivers
using these two interfaces need to be updated to use the
new pci_enable_msi_range()  or pci_enable_msi_exact()
and pci_enable_msix_range() or pci_enable_msix_exact()
interfaces.
Signed-off-by: NAlexander Gordeev <agordeev@redhat.com>
Acked-by: N"Suma Ramars (sramars)" <sramars@cisco.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

 - add host_lock option whose default value is 0 which removes the
   host_lock around all queued commands
 - accept delay=-1 (_hi_) or -2 which use a tasklet to invoke
   the scsi_done callback into the mid-layer. The default
   is still delay=1 which uses a timer to delay 1 jiffy
 - wire .change_queue_depth and .change_queue_type
   functions to better simulate queueing in a modern LLD
 - add SCSI_DEBUG_OPT_Q_NOISE (0x200) mask to only produce
   debug output associated with queue full, plus from
   .change_queue_depth and .change_queue_type functions
 - add SCSI_DEBUG_OPT_ALL_TSF (0x400) mask which reports
   all queued_arr fulls at TASK_SET_FULL, otherwise
   SCSI_MLQUEUE_HOST_BUSY is returned
 - add SCSI_DEBUG_OPT_RARE_TSF (0x800) mask which works
   together with the every_nth option (> 0) to count
   occurrences of num_in_q==queue_depth. When every_nth
   is reached the victim (a command) yields TASK SET FULL
 - clean up many debug messages.
 - add ndelay=<nanosecs> option that uses high resolution
   timers; active if > 0 and then overrides delay= option
 - expand Unit Attention handling: POR, BUS_RESET and
   MODE PARAMETERS CHANGED
 - support .eh_target_reset_handler and drop .bios_param
 - add OPT_N_WCE mask so caching page yields WCE=0
 - add OPT_RESET_NOISE mask to log aborts and resets
 - add OPT_NO_CDB_NOISE mask to not log each cdb
 - MODE SELECT support for changing caching page's WCE
 - name common ioctls in log
 - when fake_rw=1, do not vmalloc fake store; make
   UNMAP and WRITE SAME obey fake_rw
 - more logging and code improvements including better
   sense buffer handling

With fio and four (pseudo) devices I have observed 1.2 M IOPS
on my equipment. Rob Elliott who has done much testing and made
numerous suggestions, has better IOPS results than mine.
Signed-off-by: NDouglas Gilbert <dgilbert@interlog.com>
Reviewed-by: NRobert Elliott <elliott@hp.com>
Tested-by: NRobert Elliott <elliott@hp.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Add blist flags to permit the reading of the VPD pages even when
the target may claim SPC-2 compliance. MSFT targets currently
claim SPC-2 compliance while they implement post SPC-2 features.
With this patch we can correctly handle WRITE_SAME_16 issues.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Fix many warnings with incorrect endian assumptions
which makes the code unportable to new architectures.

The UFS specification defines the byte order as big-endian
for UPIU structure and little-endian for the host controller
transfer/task management descriptors.
Signed-off-by: NSujit Reddy Thumma <sthumma@codeaurora.org>
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Make undeclared functions static to suppress warnings
from sparse tool.
Signed-off-by: NSujit Reddy Thumma <sthumma@codeaurora.org>
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Signed-off-by: NVikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Instead of copying information to fw_control_context free it.

The task is forgotten thus also the reference to fw_control_context
and the completion thread takes the info from virt_ptr again.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The driver checks the return valu, but after he tries to wait_for_completion
which might never happen.  Also the ioctl buffer is freed at the end of the
function, so the first removal is not needed.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

ccb->fw_control_context is copied to local fw_control_context and
the local variable is never used later

Free ccb->fw_control_context.  The task is forgotten thus also the
reference to fw_control_context and the completion thread takes the info
from virt_ptr again.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

There is a risk that the variable will be used without being initialized.

This was largely found by using a static code analysis program called cppche
Signed-off-by: NRickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Removal of null pointer checks that could never happen
Signed-off-by: NRickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The driver ignores the return value in a lot of places, fix
it at least somewhere (and release the resources in such cases),
to avoid that bad things happen.
A memory leak is fixed too.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Patch adds a new spinlock to protect the ccb management.
It may happen that concurrent threads become the same tag value
from the 'alloc' function', the spinlock prevents this situation.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The driver ignores the return value in a lot of places, fix
it at least somewhere (and release the resources in such cases),
to avoid that bad things happen.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

In the driver two different functions are used to free the same resource,
this patch makes the code easier to read. In addittion to that, some
minor optimisations were made too.
Signed-off-by: NTomas Henzl <thenzl@redhat.com>
Acked-by: NSuresh Thiagarajan <Suresh.Thiagarajan@pmcs.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

During hibernation, the HBA firmware may lose power and forget the device
id info.   This causes the HBA to reject IO upon resume.   The fix is
to call the libsas power management routines to make the domain device
forgetful.

This fixes bug 76681: https://bugzilla.kernel.org/show_bug.cgi?id=76681Signed-off-by: NBradley Grove <bgrove@attotech.com>
Acked-by: NJack Wang <xjtuwjp@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

In _scsih_probe, delay the call to scsi_add_host until the host has been
fully set up.

Otherwise, the default .can_queue value of 1 causes scsi-mq to set the block
layer request queue size to its minimum size, resulting in awful performance.

In _scsih_probe error handling, call mpt3sas_base_detach rather than
scsi_remove_host to properly clean up in reverse order.

In _scsih_remove, call scsi_remove_host earlier to clean up in reverse order.
Signed-off-by: NRobert Elliott <elliott@hp.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NNagalakshmi Nandigama <Nagalakshmi.Nandigama@avagotech.com>
Signed-off-by: NSreekanth Reddy <Sreekanth.Reddy@avagotech.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

In _scsih_probe, delay the call to scsi_add_host until the host has been
fully set up.

Otherwise, the default .can_queue value of 1 causes scsi-mq to set the block
layer request queue size to its minimum size, resulting in awful performance.

In _scsih_probe error handling, call mpt3sas_base_detach rather than
scsi_remove_host to properly clean up in reverse order.

In _scsih_remove, call scsi_remove_host earlier to clean up in reverse order.
Signed-off-by: NRobert Elliott <elliott@hp.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NNagalakshmi Nandigama <Nagalakshmi.Nandigama@avagotech.com>
Signed-off-by: NSreekanth Reddy <Sreekanth.Reddy@avagotech.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

On some Windows hosts on FC SANs, TEST_UNIT_READY can return SRB_STATUS_ERROR.
Correctly handle this. Note that there is sufficient sense information to
support scsi error handling even in this case.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Correctly set SRB flags for all valid I/O directions. Some IHV drivers on the
Windows host require this. The host validates the command and SRB flags
prior to passing the command down to native driver stack.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

On Azure, we have seen instances of unbounded I/O latencies. To deal with
this issue, implement handler that can reset the timeout. Note that the
host gaurantees that it will respond to each command that has been issued.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
[hch: added a better comment explaining the issue]
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Based on the negotiated VMBUS protocol version, we adjust the size of the storage
protocol messages. The two sizes we currently handle are pre-win8 and post-win8.
In WS2012 R2, we are negotiating higher VMBUS protocol version than the win8
version. Make adjustments to correctly handle this.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Going forward it is possible that some of the commands that are not currently
implemented will be implemented on future Windows hosts. Even if they are not
implemented, we are told the host will corrrectly handle unsupported
commands (by returning appropriate return code and sense information).
Make command filtering depend on the host version.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Set cmd_per_lun to reflect value supported by the Host.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Hyper-V hosts can support multiple targets and multiple channels and larger number of
LUNs per target. Update the code to reflect this. With this patch we can correctly
enumerate all the paths in a multi-path storage environment.
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

If the controller doesn't support 64-bit addressing mode, it must not
set the DMA mask to 64-bit.  But it's unconditionally trying to set to
64-bit without checking 64-bit addressing support in the controller
capabilities.

It was correctly checked before commit 3b1d0580
("[SCSI] ufs: Segregate PCI Specific Code"), this aims to restores
the correct behaviour.

To achieve this in a generic way, firstly we should push down the DMA
mask setting routine ufshcd_set_dma_mask() from PCI glue driver to core
driver in order to do it for both PCI glue driver and Platform glue
driver.  Secondly, we should change pci_ DMA mapping API to dma_ DMA
mapping API because core driver is independent of glue drivers.
Signed-off-by: NAkinobu Mita <mita@fixstars.com>
Acked-by: NSantosh Y <santoshsy@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The data byte count field of PRDT indicates the length of data block
which is a segment of data transfer for SCSI commands.  The value of
this field shall have Dword granularity and the the maximum of length
is 256KB.

This adjusts dma pad mask and max segment size to the above-mentioned
PRDT limitations.
Signed-off-by: NAkinobu Mita <mita@fixstars.com>
Reviewed-by: NSubhash Jadavani <subhashj@codeaurora.org>
Tested-by: NDolev Raviv <draviv@codeaurora.org>
Acked-by: NVinayak Holikatti <vinholikatti@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

UFS 1.1 specification does not support MAINTENANCE IN(0xA3) SCSI
command and hence it doesn't support REPORT SUPPORTED OPERATION CODES
as well.

Change-Id: Ic09c5b46b2511b1c28db478023c32b898ac69e6d
Signed-off-by: NSujit Reddy Thumma <sthumma@codeaurora.org>
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

In interrupt context, after reading and comparing the UTRLDBR to
hba->outstanding_request and before resetting the interrupt aggregation,
there might be completion of another transfer request (TR). Such TRs might
get stuck, pending, until the next interrupt is generated (if any).
Changing the sequence of resetting the interrupt aggregation first and
then reading UTRLDBR status, will assure that completed TRs won't get
stuck pending.
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Some UFS devices may expose bLUQueueDepth field as zero indicating
that the queue depth depends on the number of resources available
for LUN at a particular instant to handle the outstanding transfer
requests. Currently, when response for SCSI command is TASK_FULL
the LLD decrements the queue depth but fails to increment when the
resources are available. The scsi mid-layer handles the change in
queue depth heuristically and offers simple interface with
->change_queue_depth.
Signed-off-by: NSujit Reddy Thumma <sthumma@codeaurora.org>
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Acked-by: NSantosh Y <santoshsy@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Some of the UFS devices may support different number of commands
that can be queued per LU. At the current implementation,
SW configure each of the UFS devices LU's according to the
controller capability.

In this patch the queue depth available per LU is read and updated in
the LU's SW structure.
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NRaviv Shvili <rshvili@codeaurora.org>
Acked-by: NSantosh Y <santoshsy@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Check query response status before copying the response.
Add descriptor query response size check, before copying it to buffer.
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NRaviv Shvili <rshvili@codeaurora.org>
Acked-by: NSantosh Y <santoshsy@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

Introduces the API for sending queries with descriptors.
A descriptor is a block or page of parameters that describe the device.
The descriptors are classified into types and can range in size
from 2 bytes through 255 bytes.
All descriptors have a length value as their first element, and a type
identification element as their second byte.
All descriptors are readable and some may be write once.
They are accessed using their type, index and selector.
Signed-off-by: NDolev Raviv <draviv@codeaurora.org>
Signed-off-by: NRaviv Shvili <rshvili@codeaurora.org>
Acked-by: NSantosh Y <santoshsy@gmail.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The if_info pointer is not released by the mgmt_set_ip() function
Signed-off-by: NMaurizio Lombardi <mlombard@redhat.com>
Reviewed-by: NTomas Henzl <thenzl@redhat.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

change_queue_depth allows changing per-target queue depth via sysfs.

It also allows the SCSI midlayer to ramp down the number of concurrent
inflight requests in response to a SCSI BUSY status response and allows
the midlayer to ramp the count back up to the device maximum when the
BUSY condition has resolved.
Signed-off-by: NVenkatesh Srinivas <venkateshs@google.com>
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

The spinlock of tgt_lock is only for serializing read and write
req_vq, one lockless seqcount is enough for the purpose.

On one 16core VM with vhost-scsi backend, the patch can improve
IOPS with 3% on random read test.
Signed-off-by: NMing Lei <ming.lei@canonical.com>
[Add initialization in virtscsi_target_alloc. - Paolo]
Signed-off-by: NPaolo Bonzini <pbonzini@redhat.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

When copy_from_user fails, return -EFAULT, not -ENOMEM
Signed-off-by: NStephen M. Cameron <scameron@beardog.cce.hp.com>
Reported-by: NRobert Elliott <elliott@hp.com>
Reviewed-by: NJoe Handzik <joseph.t.handzik@hp.com>
Reviewed-by: NScott Teel <scott.teel@hp.com>
Reviewed by: Mike MIller <michael.miller@canonical.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

When devices come on line, they should be removed from the list of
offline devices that are monitored.
Signed-off-by: NStephen M. Cameron <scameron@beardog.cce.hp.com>
Reviewed-by: NScott Teel <scott.teel@hp.com>
Reviewed-by: NJoe Handzik <joseph.t.handzik@hp.com>
Reviewed by: Mike MIller <michael.miller@canonical.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>

commit 28e13446 "[SCSI] hpsa: enable unit attention reporting"
turns on unit attention notifications, but got the change wrong for
all architectures other than x86, which now store an uninitialized
value into the device register.

Gcc helpfully warns about this:

../drivers/scsi/hpsa.c: In function 'hpsa_set_driver_support_bits':
../drivers/scsi/hpsa.c:6373:17: warning: 'driver_support' is used uninitialized in this function [-Wuninitialized]
  driver_support |= ENABLE_UNIT_ATTN;
                 ^

This moves the #ifdef so only the prefetch-enable is conditional
on x86, not also reading the initial register contents.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Fixes: 28e13446 "[SCSI] hpsa: enable unit attention reporting"
Cc: stable@vger.kernel.org # v3.14+
Acked-by: NStephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>