scsi: core: Fix handling of SCMD_FAIL_IF_RECOVERING

There is code in the SCSI core that sets the SCMD_FAIL_IF_RECOVERING
flag but there is no code that clears this flag. Instead of only clearing
SCMD_INITIALIZED in scsi_end_request(), clear all flags. It is never
necessary to preserve any command flags inside scsi_end_request().

Cc: stable@vger.kernel.org
Fixes: 310bcaef6d7e ("scsi: core: Support failing requests while recovering")
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20240325224417.1477135-1-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Really include kunit tests with SCSI_LIB_KUNIT_TEST

Commit 25a1f7a0a1fe ("scsi: core: Add kunit tests for
scsi_check_passthrough()") adds the config SCSI_LIB_KUNIT_TEST and
corresponding tests. Due to naming confusion, the actual tests would only
be included when the non-existing config SCSI_KUNIT_TEST is enabled (note
this missing 'LIB' in the config name). So, they are basically dead right
now.

Adjust the name to actual existing config.

Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Link: https://lore.kernel.org/r/20240207145603.15680-1-lukas.bulwahn@gmail.com
Reviewed-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add kunit tests for scsi_check_passthrough()

Add some kunit tests for scsi_check_passthrough() so we can easily make
sure we are hitting the cases for which it's difficult to replicate in
hardware or even scsi_debug.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/20240123002220.129141-20-michael.christie@oracle.com
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Have midlayer retry scsi_mode_sense() UAs

This has scsi_mode_sense() have the SCSI midlayer retry UAs instead of
driving them itself.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/20240123002220.129141-13-michael.christie@oracle.com
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Allow passthrough to request midlayer retries

For passthrough we don't retry any error which we get a check condition
for. This results in a lot of callers driving their own retries for all
UAs, specific UAs, NOT_READY, specific sense values or any type of failure.

This adds the core code to allow passthrough users to specify what errors
they want the SCSI midlayer to retry for them. We can then convert users to
drop a lot of their sense parsing and retry handling.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/20240123002220.129141-2-michael.christie@oracle.com
Reviewed-by: John Garry <john.g.garry@oracle.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Move scsi_host_busy() out of host lock if it is for per-command

Commit 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock
for waking up EH handler") intended to fix a hard lockup issue triggered by
EH. The core idea was to move scsi_host_busy() out of the host lock when
processing individual commands for EH. However, a suggested style change
inadvertently caused scsi_host_busy() to remain under the host lock. Fix
this by calling scsi_host_busy() outside the lock.

Fixes: 4373534a9850 ("scsi: core: Move scsi_host_busy() out of host lock for waking up EH handler")
Cc: Sathya Prakash Veerichetty <safhya.prakash@broadcom.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20240203024521.2006455-1-ming.lei@redhat.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Move scsi_host_busy() out of host lock for waking up EH handler

Inside scsi_eh_wakeup(), scsi_host_busy() is called & checked with host
lock every time for deciding if error handler kthread needs to be waken up.

This can be too heavy in case of recovery, such as:

- N hardware queues

- queue depth is M for each hardware queue

- each scsi_host_busy() iterates over (N * M) tag/requests

If recovery is triggered in case that all requests are in-flight, each
scsi_eh_wakeup() is strictly serialized, when scsi_eh_wakeup() is called
for the last in-flight request, scsi_host_busy() has been run for (N * M -
1) times, and request has been iterated for (N*M - 1) * (N * M) times.

If both N and M are big enough, hard lockup can be triggered on acquiring
host lock, and it is observed on mpi3mr(128 hw queues, queue depth 8169).

Fix the issue by calling scsi_host_busy() outside the host lock. We don't
need the host lock for getting busy count because host the lock never
covers that.

[mkp: Drop unnecessary 'busy' variables pointed out by Bart]

Cc: Ewan Milne <emilne@redhat.com>
Fixes: 6eb045e092ef ("scsi: core: avoid host-wide host_busy counter for scsi_mq")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/20240112070000.4161982-1-ming.lei@redhat.com
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: Sathya Prakash Veerichetty <safhya.prakash@broadcom.com>
Tested-by: Sathya Prakash Veerichetty <safhya.prakash@broadcom.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Clean up scsi_dev_queue_ready()

This is just a cleanup for scsi_dev_queue_ready() to avoid a redundant goto
and if statement. No functional change.

Signed-off-by: Wenchao Hao <haowenchao2@huawei.com>
Link: https://lore.kernel.org/r/20231018113746.1940197-2-haowenchao2@huawei.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Handle depopulation and restoration in progress

The default handling of the NOT READY sense key is to wait for the device
to become ready. The "wait" is assumed to be relatively short. However
there is a sub-class of NOT READY that have the "... in progress" phrase in
their additional sense code and these can take much longer. Following on
from commit 505aa4b6a883 ("scsi: sd: Defer spinning up drive while SANITIZE
is in progress") we now have element depopulation and restoration that can
take a long time. For example, over 24 hours for a 20 TB, 7200 rpm hard
disk to depopulate 1 of its 20 elements.

Add handling of ASC/ASCQ: 0x4,0x24 (depopulation in progress)
and ASC/ASCQ: 0x4,0x25 (depopulation restoration in progress)
to sd.c . The scsi_lib.c has incomplete handling of these
two messages, so complete it.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Link: https://lore.kernel.org/r/20231015050650.131145-1-dgilbert@interlog.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Fix sshdr use in scsi_test_unit_ready

If scsi_execute_cmd returns < 0, it doesn't initialize the sshdr, so we
shouldn't access the sshdr. If it returns 0, then the cmd executed
successfully, so there is no need to check the sshdr. This has us access
the sshdr when we get a return value > 0.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/20231004210013.5601-10-michael.christie@oracle.com
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Improve type safety of scsi_rescan_device()

Most callers of scsi_rescan_device() have the scsi_device pointer readily
available. Pass a struct scsi_device pointer to scsi_rescan_device()
instead of a struct device pointer. This change prevents that a pointer to
another struct device would be passed accidentally to scsi_rescan_device().

Remove the scsi_rescan_device() declaration from the scsi_priv.h header
file since it duplicates the declaration in <scsi/scsi_host.h>.

Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Cc: Mike Christie <michael.christie@oracle.com>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230822153043.4046244-1-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Improve performance for BLK_MQ_F_BLOCKING drivers

blk_mq_run_queue() runs the queue asynchronously if BLK_MQ_F_BLOCKING
has been set. This is suboptimal since running the queue asynchronously
is slower than running the queue synchronously. This patch modifies
blk_mq_run_queue() as follows if BLK_MQ_F_BLOCKING has been set:
- Run the queue synchronously if it is allowed to sleep.
- Run the queue asynchronously if it is not allowed to sleep.
Additionally, blk_mq_run_hw_queue(hctx, false) calls are modified into
blk_mq_run_hw_queue(hctx, hctx->flags & BLK_MQ_F_BLOCKING) if the caller
may be invoked from atomic context.

The following caller chains have been reviewed:

blk_mq_run_hw_queue(hctx, false)
blk_mq_get_tag() /* may sleep, hence the functions it calls may also sleep */
blk_execute_rq() /* may sleep */
blk_mq_run_hw_queues(q, async=false)
blk_freeze_queue_start() /* may sleep */
blk_mq_requeue_work() /* may sleep */
scsi_kick_queue()
scsi_requeue_run_queue() /* may sleep */
scsi_run_host_queues()
scsi_ioctl_reset() /* may sleep */
blk_mq_insert_requests(hctx, ctx, list, run_queue_async=false)
blk_mq_dispatch_plug_list(plug, from_sched=false)
blk_mq_flush_plug_list(plug, from_schedule=false)
__blk_flush_plug(plug, from_schedule=false)
blk_add_rq_to_plug()
blk_mq_submit_bio() /* may sleep if REQ_NOWAIT has not been set */
blk_mq_plug_issue_direct()
blk_mq_flush_plug_list() /* see above */
blk_mq_dispatch_plug_list(plug, from_sched=false)
blk_mq_flush_plug_list() /* see above */
blk_mq_try_issue_directly()
blk_mq_submit_bio() /* may sleep if REQ_NOWAIT has not been set */
blk_mq_try_issue_list_directly(hctx, list)
blk_mq_insert_requests() /* see above */

Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230721172731.955724-4-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: Remove a blk_mq_run_hw_queues() call

blk_mq_kick_requeue_list() calls blk_mq_run_hw_queues() asynchronously.
Leave out the direct blk_mq_run_hw_queues() call. This patch causes
scsi_run_queue() to call blk_mq_run_hw_queues() asynchronously instead
of synchronously. Since scsi_run_queue() is not called from the hot I/O
submission path, this patch does not affect the hot path.

This patch prepares for allowing blk_mq_run_hw_queue() to sleep if
BLK_MQ_F_BLOCKING has been set. scsi_run_queue() may be called from
atomic context and must not sleep. Hence the removal of the
blk_mq_run_hw_queues(q, false) call. See also scsi_unblock_requests().

Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: "Martin K. Petersen" <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20230721172731.955724-3-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: Inline scsi_kick_queue()

Inline scsi_kick_queue() to prepare for modifying the second argument
passed to blk_mq_run_hw_queues().

Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: "Martin K. Petersen" <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20230721172731.955724-2-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Improve warning message in scsi_device_block()

If __scsi_internal_device_block() returns an error, it is always -EINVAL
because of an invalid state transition. For debugging purposes, it makes
more sense to print the device state.

Signed-off-by: Martin Wilck <mwilck@suse.com>
Link: https://lore.kernel.org/r/20230614103616.31857-8-mwilck@suse.com
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Replace scsi_target_block() with scsi_block_targets()

All callers (fc_remote_port_delete(), __iscsi_block_session(),
__srp_start_tl_fail_timers(), srp_reconnect_rport(), snic_tgt_del()) pass
parent devices of scsi_target devices to scsi_target_block().

Rename the function to scsi_block_targets(), and simplify it by assuming
that it is always passed a parent device. Also, have callers pass the
Scsi_Host pointer to scsi_block_targets(), as every caller has this pointer
readily available.

Suggested-by: Christoph Hellwig <hch@lst.de>
Suggested-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin Wilck <mwilck@suse.com>
Link: https://lore.kernel.org/r/20230614103616.31857-7-mwilck@suse.com
Cc: Karan Tilak Kumar <kartilak@cisco.com>
Cc: Sesidhar Baddela <sebaddel@cisco.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Don't wait for quiesce in scsi_device_block()

scsi_device_block() is only called from scsi_target_block(), which calls it
repeatedly for every child device. For targets with many devices, waiting
for every queue to quiesce may cause a substantial delay (we measured more
than 100s delay for blocking a FC rport with 2048 LUNs).

Just call blk_mq_wait_quiesce_done() once from scsi_target_block() after
stopping all queues.

Signed-off-by: Martin Wilck <mwilck@suse.com>
Link: https://lore.kernel.org/r/20230614103616.31857-6-mwilck@suse.com
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Don't wait for quiesce in scsi_stop_queue()

scsi_stop_queue() has just two callers, one with and one without
"nowait". As blk_mq_quiesce_queue() comes down to
blk_mq_quiesce_queue_nowait() followed by blk_mq_wait_quiesce_done(), we
might as well open-code this in scsi_device_block().

Also, add a comment explaining why blk_mq_quiesce_queue_nowait() must be
called with the state_mutex held, see
https://lore.kernel.org/linux-scsi/3b8b13bf-a458-827a-b916-07d7eee8ae00@acm.org/.

Signed-off-by: Martin Wilck <mwilck@suse.com>
Link: https://lore.kernel.org/r/20230614103616.31857-5-mwilck@suse.com
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Merge scsi_internal_device_block() and device_block()

scsi_internal_device_block() is only called from device_block(). Merge the
two functions, and call the result scsi_device_block(), as the name
device_block() is confusingly generic.

Signed-off-by: Martin Wilck <mwilck@suse.com>
Link: https://lore.kernel.org/r/20230614103616.31857-4-mwilck@suse.com
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Support setting BLK_MQ_F_BLOCKING

Prepare for adding code in ufshcd_queuecommand() that may sleep. This patch
is similar to a patch posted last year by Mike Christie. See also
https://lore.kernel.org/all/20220308003957.123312-2-michael.christie@oracle.com/

Cc: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230529202640.11883-3-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rework scsi_host_block()

Make scsi_host_block() easier to read by converting it to the widely used
early-return style. See also commit f983622ae605 ("scsi: core: Avoid
calling synchronize_rcu() for each device in scsi_host_block()").

Reviewed-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Cc: Ye Bin <yebin10@huawei.com>
Cc: Hannes Reinecke <hare@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230529202640.11883-2-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Only kick the requeue list if necessary

Instead of running the request queue of each device associated with a host
every 3 ms (BLK_MQ_RESOURCE_DELAY) while host error handling is in
progress, run the request queue after error handling has finished.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.g.garry@oracle.com>
Cc: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20230518193159.1166304-4-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: sd: Handle read/write CDL timeout failures

Commands using a duration limit descriptor that has limit policies set to a
value other than 0x0 may be failed by the device if one of the limits are
exceeded. For such commands, since the failure is the result of the user
duration limit configuration and workload, the commands should not be
retried and terminated immediately. Furthermore, to allow the user to
differentiate these "soft" failures from hard errors due to hardware
problem, a different error code than EIO should be returned.

There are 2 cases to consider:

(1) The failure is due to a limit policy failing the command with a check
condition sense key, that is, any limit policy other than 0xD. For this
case, scsi_check_sense() is modified to detect failures with the ABORTED
COMMAND sense key and the COMMAND TIMEOUT BEFORE PROCESSING or COMMAND
TIMEOUT DURING PROCESSING or COMMAND TIMEOUT DURING PROCESSING DUE TO ERROR
RECOVERY additional sense code. For these failures, a SUCCESS disposition
is returned so that scsi_finish_command() is called to terminate the
command.

(2) The failure is due to a limit policy set to 0xD, which result in the
command being terminated with a GOOD status, COMPLETED sense key, and DATA
CURRENTLY UNAVAILABLE additional sense code. To handle this case, the
scsi_check_sense() is modified to return a SUCCESS disposition so that
scsi_finish_command() is called to terminate the command. In addition,
scsi_decide_disposition() has to be modified to see if a command being
terminated with GOOD status has sense data. This is as defined in SCSI
Primary Commands - 6 (SPC-6), so all according to spec, even if GOOD status
commands were not checked before.

If scsi_check_sense() detects sense data representing a duration limit,
scsi_check_sense() will set the newly introduced SCSI ML byte
SCSIML_STAT_DL_TIMEOUT. This SCSI ML byte is checked in scsi_noretry_cmd(),
so that a command that failed because of a CDL timeout cannot be
retried. The SCSI ML byte is also checked in scsi_result_to_blk_status() to
complete the command request with the BLK_STS_DURATION_LIMIT status, which
result in the user seeing ETIME errors for the failed commands.

Co-developed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Link: https://lore.kernel.org/r/20230511011356.227789-12-nks@flawful.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Support retrieving sub-pages of mode pages

Allow scsi_mode_sense() to retrieve sub-pages of mode pages by adding the
subpage argument. Change all the current caller sites to specify the
subpage 0.

Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Link: https://lore.kernel.org/r/20230511011356.227789-7-nks@flawful.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename and move get_scsi_ml_byte()

SCSI has two different getters:

- get_XXX_byte() (in scsi_cmnd.h) which takes a struct scsi_cmnd *, and

- XXX_byte() (in scsi.h) which takes a scmd->result.

The proper name for get_scsi_ml_byte() should thus be without the get_
prefix, as it takes a scmd->result. Rename the function to rectify this.
(This change was suggested by Mike Christie.)

Additionally, move get_scsi_ml_byte() to scsi_priv.h since both scsi_lib.c
and scsi_error.c will need to use this helper in a follow-up patch.

Cc: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Link: https://lore.kernel.org/r/20230511011356.227789-6-nks@flawful.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Rename BLK_STS_NEXUS to BLK_STS_RESV_CONFLICT

BLK_STS_NEXUS is used for NVMe/SCSI reservation conflicts and DASD's
locking feature which works similar to NVMe/SCSI reservations where a
host can get a lock on a device and when the lock is taken it will get
failures.

This patch renames BLK_STS_NEXUS so it better reflects this type of
use.

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Link: https://lore.kernel.org/r/20230407200551.12660-3-michael.christie@oracle.com
Acked-by: Stefan Haberland <sth@linux.ibm.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Decrease scsi_device's iorequest_cnt if dispatch failed

If scsi_dispatch_cmd() failed, the SCSI command was not sent to the target,
scsi_queue_rq() would return BLK_STS_RESOURCE and the related request would
be requeued. The timeout of this request would not fire, no one would
increase iodone_cnt.

The above flow would result the iodone_cnt smaller than iorequest_cnt. So
decrease the iorequest_cnt if dispatch failed to workaround the issue.

Signed-off-by: Wenchao Hao <haowenchao2@huawei.com>
Reported-by: Ming Lei <ming.lei@redhat.com>
Closes: https://lore.kernel.org/r/ZF+zB+bB7iqe0wGd@ovpn-8-17.pek2.redhat.com
Link: https://lore.kernel.org/r/20230515070156.1790181-3-haowenchao2@huawei.com
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Revert "scsi: core: Do not increase scsi_device's iorequest_cnt if dispatch failed"

The "atomic_inc(&cmd->device->iorequest_cnt)" in scsi_queue_rq() would
cause kernel panic because cmd->device may be freed after returning from
scsi_dispatch_cmd().

This reverts commit cfee29ffb45b1c9798011b19d454637d1b0fe87d.

Signed-off-by: Wenchao Hao <haowenchao2@huawei.com>
Reported-by: Ming Lei <ming.lei@redhat.com>
Closes: https://lore.kernel.org/r/ZF+zB+bB7iqe0wGd@ovpn-8-17.pek2.redhat.com
Link: https://lore.kernel.org/r/20230515070156.1790181-2-haowenchao2@huawei.com
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Extend struct scsi_exec_args

Allow SCSI LLDs to specify SCMD_* flags.

Link: https://lore.kernel.org/r/20230210193258.4004923-2-bvanassche@acm.org
Cc: Mike Christie <michael.christie@oracle.com>
Cc: John Garry <john.g.garry@oracle.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Convert to scsi_execute_cmd()

scsi_execute_req() is going to be removed. Convert SCSI midlayer to
scsi_execute_cmd().

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add struct for args to execution functions

Move the SCSI execution functions to use a struct for passing in optional
args. This commit adds the new struct, temporarily converts scsi_execute()
and scsi_execute_req() ands a new helper, scsi_execute_cmd(), which takes
the scsi_exec_args struct.

There should be no change in behavior. We no longer allow users to pass in
any request->rq_flags value, but they were only passing in RQF_PM which we
do support by allowing users to pass in the BLK_MQ_REQ flags used by
blk_mq_alloc_request().

Subsequent commits will convert scsi_execute() and scsi_execute_req() users
to the new helpers then remove scsi_execute() and scsi_execute_req().

Signed-off-by: Mike Christie <michael.christie@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Do not increase scsi_device's iorequest_cnt if dispatch failed

If scsi_dispatch_cmd() failed, the SCSI command was not sent to the target.
scsi_queue_rq() would return BLK_STS_RESOURCE if scsi_dispatch_cmd()
failed, and the related request would be requeued. The timeout of this
request would not fire, so noone would increase iodone_cnt.

Signed-off-by: Wenchao Hao <haowenchao@huawei.com>
Link: https://lore.kernel.org/r/20221123122137.150776-3-haowenchao@huawei.com
Reviewed-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Support failing requests while recovering

The current behavior for SCSI commands submitted while error recovery is
ongoing is to retry command submission after error recovery has finished.
See also the scsi_host_in_recovery() check in scsi_host_queue_ready(). Add
support for failing SCSI commands while host recovery is in progress. This
functionality will be used to fix a deadlock in the UFS driver.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: John Garry <john.garry@huawei.com>
Cc: Mike Christie <michael.christie@oracle.com>
Cc: Hannes Reinecke <hare@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20221018202958.1902564-4-bvanassche@acm.org
Reviewed-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rework scsi_single_lun_run()

Use __starget_for_each_device() instead of open-coding
starget_for_each_device(). Run the queues asynchronously instead of
synchronously.

This commit removes code that calls scsi_device_put() from atomic context.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.garry@huawei.com>
Cc: Mike Christie <michael.christie@oracle.com>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20221015002418.30955-6-bvanassche@acm.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

blk-mq: pass a tagset to blk_mq_wait_quiesce_done

Nothing in blk_mq_wait_quiesce_done needs the request_queue now, so just
pass the tagset, and move the non-mq check into the only caller that
needs it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chao Leng <lengchao@huawei.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20221101150050.3510-13-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Convert scsi_decide_disposition() to use SCSIML_STAT

Don't use:

- DID_TARGET_FAILURE

- DID_NEXUS_FAILURE

- DID_ALLOC_FAILURE

- DID_MEDIUM_ERROR

Instead use the SCSI midlayer internal values.

Link: https://lore.kernel.org/r/20220812010027.8251-10-michael.christie@oracle.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add error codes for internal SCSI midlayer use

If a driver returns:

- DID_TARGET_FAILURE

- DID_NEXUS_FAILURE

- DID_ALLOC_FAILURE

- DID_MEDIUM_ERROR

we hit a couple bugs:

1. The SCSI error handler runs because scsi_decide_disposition() has no
case statements for them and we return FAILED.

2. For SG IO the userspace app gets a success status instead of failed,
because scsi_result_to_blk_status() clears those errors.

This patch adds a new internal error code byte for use by the SCSI
midlayer. This will be used instead of the above error codes, so we don't
have to play that clearing the host code game in
scsi_result_to_blk_status() and drivers cannot accidentally use them.

A subsequent commit will then remove the internal users of the above codes
and convert us to use the new ones.

Link: https://lore.kernel.org/r/20220812010027.8251-9-michael.christie@oracle.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Change the return type of blk_mq_map_queues() into void

Since blk_mq_map_queues() and the .map_queues() callbacks always return 0,
change their return type into void. Most callers ignore the returned value
anyway.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Keith Busch <kbusch@kernel.org>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Doug Gilbert <dgilbert@interlog.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Acked-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Link: https://lore.kernel.org/r/20220815170043.19489-3-bvanassche@acm.org
[axboe: fold in fix from Bart]
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Fix a use-after-free

There are two .exit_cmd_priv implementations. Both implementations use
resources associated with the SCSI host. Make sure that these resources are
still available when .exit_cmd_priv is called by waiting inside
scsi_remove_host() until the tag set has been freed.

This commit fixes the following use-after-free:

==================================================================
BUG: KASAN: use-after-free in srp_exit_cmd_priv+0x27/0xd0 [ib_srp]
Read of size 8 at addr ffff888100337000 by task multipathd/16727
Call Trace:
<TASK>
dump_stack_lvl+0x34/0x44
print_report.cold+0x5e/0x5db
kasan_report+0xab/0x120
srp_exit_cmd_priv+0x27/0xd0 [ib_srp]
scsi_mq_exit_request+0x4d/0x70
blk_mq_free_rqs+0x143/0x410
__blk_mq_free_map_and_rqs+0x6e/0x100
blk_mq_free_tag_set+0x2b/0x160
scsi_host_dev_release+0xf3/0x1a0
device_release+0x54/0xe0
kobject_put+0xa5/0x120
device_release+0x54/0xe0
kobject_put+0xa5/0x120
scsi_device_dev_release_usercontext+0x4c1/0x4e0
execute_in_process_context+0x23/0x90
device_release+0x54/0xe0
kobject_put+0xa5/0x120
scsi_disk_release+0x3f/0x50
device_release+0x54/0xe0
kobject_put+0xa5/0x120
disk_release+0x17f/0x1b0
device_release+0x54/0xe0
kobject_put+0xa5/0x120
dm_put_table_device+0xa3/0x160 [dm_mod]
dm_put_device+0xd0/0x140 [dm_mod]
free_priority_group+0xd8/0x110 [dm_multipath]
free_multipath+0x94/0xe0 [dm_multipath]
dm_table_destroy+0xa2/0x1e0 [dm_mod]
__dm_destroy+0x196/0x350 [dm_mod]
dev_remove+0x10c/0x160 [dm_mod]
ctl_ioctl+0x2c2/0x590 [dm_mod]
dm_ctl_ioctl+0x5/0x10 [dm_mod]
__x64_sys_ioctl+0xb4/0xf0
dm_ctl_ioctl+0x5/0x10 [dm_mod]
__x64_sys_ioctl+0xb4/0xf0
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x46/0xb0

Link: https://lore.kernel.org/r/20220826002635.919423-1-bvanassche@acm.org
Fixes: 65ca846a5314 ("scsi: core: Introduce {init,exit}_cmd_priv()")
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Mike Christie <michael.christie@oracle.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.garry@huawei.com>
Cc: Li Zhijian <lizhijian@fujitsu.com>
Reported-by: Li Zhijian <lizhijian@fujitsu.com>
Tested-by: Li Zhijian <lizhijian@fujitsu.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix passthrough retry counter handling

Passthrough users will set the scsi_cmnd->allowed value and were expecting
up to $allowed retries. The problem is that before:

commit 6aded12b10e0 ("scsi: core: Remove struct scsi_request")

we used to set the retries on the scsi_request then copy them over to
scsi_cmnd->allowed in scsi_setup_scsi_cmnd. With that patch we now set
scsi_cmnd->allowed to 0 in scsi_prepare_cmd and overwrite what the
passthrough user set.

This moves the allowed initialization to after the blk_rq_is_passthrough()
check so it's only done for the non-passthrough path where the ULD
init_command will normally set an allowed value it prefers.

Link: https://lore.kernel.org/r/20220812011206.9157-1-michael.christie@oracle.com
Fixes: 6aded12b10e0 ("scsi: core: Remove struct scsi_request")
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Allow the ALUA transitioning state enough time

The error path for the SCSI check condition of not ready, target in ALUA
state transition, will result in the failure of that path after the retries
are exhausted. In most cases that is well ahead of the transition timeout
established in the SCSI ALUA device handler.

Instead, reprep the command and re-add it to the queue after a 1 second
delay. This will allow the handler to take care of the timeout and only
fail the path if the target has exceeded the transition expiry timeout
(default 60 seconds). If the expiry timeout is exceeded, the handler will
change the path state from transitioning to standby leading to a path
failure eliminating the potential of this re-prep to continue endlessly. In
most cases the target will exit the transitioning state well before the
expiry timeout but after the retries are exhausted as mentioned.

Additionally remove the scsi_io_completion_reprep() function which provides
little value.

Link: https://lore.kernel.org/r/20220729214110.58576-1-brian@purestorage.com
Reviewed-by: Martin Wilck <mwilck@suse.com>
Acked-by: Krishna Kant <krishna.kant@purestorage.com>
Acked-by: Seamus Connor <sconnor@purestorage.com>
Signed-off-by: Brian Bunker <brian@purestorage.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: cap shost max_sectors according to DMA limits only once

The shost->max_sectors is repeatedly capped according to the host DMA
mapping limit for each sdev in __scsi_init_queue(). This is unnecessary, so
set only once when adding the host.

Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>

scsi: core: Move the definition of SCSI_QUEUE_DELAY

Move the definition of SCSI_QUEUE_DELAY to just above the function that
uses it.

Link: https://lore.kernel.org/r/20220630195703.10155-2-bvanassche@acm.org
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.garry@huawei.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi/core: Use the new blk_opf_t type

Use the new blk_opf_t type for arguments and variables that represent
request flags. Use the !! operator in scsi_noretry_cmd() to convert the
blk_opf_t type into a boolean. This patch does not change any functionality.

Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.garry@huawei.com>
Cc: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20220714180729.1065367-42-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi/core: Improve static type checking

Improve static type checking by using the new blk_opf_t type for the
combination of a request operation and its flags.

Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: John Garry <john.garry@huawei.com>
Cc: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20220714180729.1065367-40-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Remove reserved request time-out handling

The SCSI core code does not currently support reserved commands. As such,
requests which time-out would never be reserved, and scsi_timeout()
'reserved' arg should never be set.

Remove handling for reserved requests, drop the wrapper scsi_timeout()
as it now just calls scsi_times_out() always, and finally rename
scsi_times_out() -> scsi_timeout() to match the blk_mq_ops method name.

Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/1657109034-206040-2-git-send-email-john.garry@huawei.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: simplify disk shutdown

Set the queue dying flag and call blk_mq_exit_queue from del_gendisk for
all disks that do not have separately allocated queues, and thus remove
the need to call blk_cleanup_queue for them.

Rename blk_cleanup_disk to blk_mq_destroy_queue to make it clear that
this function is intended only for separately allocated blk-mq queues.

This saves an extra queue freeze for devices without a separately
allocated queue.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20220619060552.1850436-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Return BLK_STS_TRANSPORT for ALUA transitioning

When the 'ALUA state transitioning' sense code is returned we cannot use
BLK_STS_AGAIN, as this has a very specific use-case. So return
BLK_STS_TRANSPORT here.

Link: https://lore.kernel.org/r/20220524055631.85480-3-hare@suse.de
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Refine how we set tag_set NUMA node

For SCSI hosts which enable host_tagset the NUMA node returned from
blk_mq_hw_queue_to_node() is NUMA_NO_NODE always. Then, since in
scsi_mq_setup_tags() the default we choose for the tag_set NUMA node is
NUMA_NO_NODE, we always evaluate the NUMA node as NUMA_NO_NODE in functions
like blk_mq_alloc_rq_map().

The reason we get NUMA_NO_NODE from blk_mq_hw_queue_to_node() is that the
hctx_idx passed is BLK_MQ_NO_HCTX_IDX - so we can't match against a (HW)
queue mapping index.

Improve this by defaulting the tag_set NUMA node to the same NUMA node of
the SCSI host DMA dev.

Link: https://lore.kernel.org/r/1648640315-21419-1-git-send-email-john.garry@huawei.com
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove struct scsi_request

Let submitters initialize the scmd->allowed field directly instead of
indirecting through struct scsi_request and remove the now superfluous
structure.

Link: https://lore.kernel.org/r/20220224175552.988286-8-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Move the result field from struct scsi_request to struct scsi_cmnd

Prepare for removing the scsi_request structure by moving the result field
to struct scsi_cmnd.

Link: https://lore.kernel.org/r/20220224175552.988286-7-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Move the resid_len field from struct scsi_request to struct scsi_cmnd

Prepare for removing the scsi_request structure by moving the resid_len
field to struct scsi_cmnd.

Link: https://lore.kernel.org/r/20220224175552.988286-6-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove the sense and sense_len fields from struct scsi_request

Just use the sense_buffer field in struct scsi_cmnd for the sense data and
move the sense_len field over to struct scsi_cmnd.

Link: https://lore.kernel.org/r/20220224175552.988286-5-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove the cmd field from struct scsi_request

Now that each scsi_request is backed by a scsi_cmnd, there is no need to
indirect the CDB storage. Change all submitters of SCSI passthrough
requests to store the CDB information directly in the scsi_cmnd, and while
doing so allocate the full 32 bytes that cover all Linux supported SCSI
hosts instead of requiring dynamic allocation for > 16 byte CDBs. On
64-bit systems this does not change the size of the scsi_cmnd at all, while
on 32-bit systems it slightly increases it for now, but that increase will
be made up by the removal of the remaining scsi_request fields.

Link: https://lore.kernel.org/r/20220224175552.988286-4-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Don't memset() the entire scsi_cmnd in scsi_init_command()

Replace the big fat memset that requires saving and restoring various
fields with just initializing those fields that need initialization.

All the clearing to 0 is moved to scsi_prepare_cmd() as scsi_ioctl_reset()
alreadly uses kzalloc() to allocate a pre-zeroed command.

This is still conservative and can probably be optimized further.

Link: https://lore.kernel.org/r/20220224175552.988286-3-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add scsi_done_direct() for immediate completion

Add scsi_done_direct() which behaves like scsi_done() except that it
invokes blk_mq_complete_request_direct() in order to complete the request.

Callers from process context can complete the request directly instead
waking ksoftirqd.

Link: https://lore.kernel.org/r/Yfw7JaszshmfYa1d@flow
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: use BLK_STS_OFFLINE for not fully online devices

The new error message for such case looks like

[ 172.809565] device offline error, dev sda, sector 3138208 ...

which will not be confused with regular I/O error (BLK_STS_IOERR).

Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20220203192827.1370270-4-song@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Fix scsi_mode_select() interface

The modepage argument is unused. Remove it.

Link: https://lore.kernel.org/r/20210929091744.706003-3-damien.lemoal@wdc.com
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: remove the gendisk argument to blk_execute_rq

Remove the gendisk aregument to blk_execute_rq and blk_execute_rq_nowait
given that it is unused now. Also convert the boolean at_head parameter
to actually use the bool type while touching the prototype.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20211126121802.2090656-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: remove the ->rq_disk field in struct request

Just use the disk attached to the request_queue instead.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20211126121802.2090656-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move blk_rq_err_bytes to scsi

blk_rq_err_bytes is only used by the scsi midlayer, so move it there.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20211117061404.331732-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Remove command size deduction from scsi_setup_scsi_cmnd()

No need to deduce command size in scsi_setup_scsi_cmnd() anymore as
appropriate checks have been added to scsi_fill_sghdr_rq() function and the
cmd_len should never be zero here. The code to do that wasn't correct
anyway, as it used uninitialized cmd->cmnd, which caused a null-ptr-deref
if the command size was zero as in the trace below. Fix this by removing
the unneeded code.

KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007]
CPU: 0 PID: 1822 Comm: repro Not tainted 5.15.0 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-4.fc34 04/01/2014
Call Trace:
blk_mq_dispatch_rq_list+0x7c7/0x12d0
__blk_mq_sched_dispatch_requests+0x244/0x380
blk_mq_sched_dispatch_requests+0xf0/0x160
__blk_mq_run_hw_queue+0xe8/0x160
__blk_mq_delay_run_hw_queue+0x252/0x5d0
blk_mq_run_hw_queue+0x1dd/0x3b0
blk_mq_sched_insert_request+0x1ff/0x3e0
blk_execute_rq_nowait+0x173/0x1e0
blk_execute_rq+0x15c/0x540
sg_io+0x97c/0x1370
scsi_ioctl+0xe16/0x28e0
sd_ioctl+0x134/0x170
blkdev_ioctl+0x362/0x6e0
block_ioctl+0xb0/0xf0
vfs_ioctl+0xa7/0xf0
do_syscall_64+0x3d/0xb0
entry_SYSCALL_64_after_hwframe+0x44/0xae
---[ end trace 8b086e334adef6d2 ]---
Kernel panic - not syncing: Fatal exception

Link: https://lore.kernel.org/r/20211103170659.22151-2-tadeusz.struk@linaro.org
Fixes: 2ceda20f0a99 ("scsi: core: Move command size detection out of the fast path")
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: James E.J. Bottomley <jejb@linux.ibm.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: <linux-scsi@vger.kernel.org>
Cc: <linux-kernel@vger.kernel.org>
Cc: <stable@vger.kernel.org> # 5.15, 5.14, 5.10
Reported-by: syzbot+5516b30f5401d4dcbcae@syzkaller.appspotmail.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Tadeusz Struk <tadeusz.struk@linaro.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Avoid leaving shost->last_reset with stale value if EH does not run

The changes to issue the abort from the scmd->abort_work instead of the EH
thread introduced a problem if eh_deadline is used. If aborting the
command(s) is successful, and there are never any scmds added to the
shost->eh_cmd_q, there is no code path which will reset the ->last_reset
value back to zero.

The effect of this is that after a successful abort with no EH thread
activity, a subsequent timeout, perhaps a long time later, might
immediately be considered past a user-set eh_deadline time, and the host
will be reset with no attempt at recovery.

Fix this by resetting ->last_reset back to zero in scmd_eh_abort_handler()
if it is determined that the EH thread will not run to do this.

Thanks to Gopinath Marappan for investigating this problem.

Link: https://lore.kernel.org/r/20211029194311.17504-2-emilne@redhat.com
Fixes: e494f6a72839 ("[SCSI] improved eh timeout handler")
Cc: stable@vger.kernel.org
Signed-off-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: make sure that request queue queiesce and unquiesce balanced

For fixing queue quiesce race between driver and block layer(elevator
switch, update nr_requests, ...), we need to support concurrent quiesce
and unquiesce, which requires the two call balanced.

It isn't easy to audit that in all scsi drivers, especially the two may
be called from different contexts, so do it in scsi core with one
per-device atomic variable to balance quiesce and unquiesce.

Reported-by: Yi Zhang <yi.zhang@redhat.com>
Fixes: e70feb8b3e68 ("blk-mq: support concurrent queue quiesce/unquiesce")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20211109071144.181581-4-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: avoid to quiesce sdev->request_queue two times

For fixing queue quiesce race between driver and block layer(elevator
switch, update nr_requests, ...), we need to support concurrent quiesce
and unquiesce, which requires the two to be balanced.

blk_mq_quiesce_queue() calls blk_mq_quiesce_queue_nowait() for updating
quiesce depth and marking the flag, then scsi_internal_device_block() calls
blk_mq_quiesce_queue_nowait() two times actually.

Fix the double quiesce and keep quiesce and unquiesce balanced.

Reported-by: Yi Zhang <yi.zhang@redhat.com>
Fixes: e70feb8b3e68 ("blk-mq: support concurrent queue quiesce/unquiesce")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Link: https://lore.kernel.org/r/20211109071144.181581-3-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Call scsi_done directly

Conditional statements are faster than indirect calls. Hence call
scsi_done() directly. Since this patch removes the last user of the
scsi_done member, also remove that data structure member.

Link: https://lore.kernel.org/r/20211007204618.2196847-11-bvanassche@acm.org
Reviewed-by: Benjamin Block <bblock@linux.ibm.com>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename scsi_mq_done() into scsi_done() and export it

Since the removal of the legacy block layer there is only one completion
function left in the SCSI core, namely scsi_mq_done(). Rename it into
scsi_done(). Export that function to allow SCSI LLDs to call it directly.

Link: https://lore.kernel.org/r/20211007202923.2174984-3-bvanassche@acm.org
Reviewed-by: Benjamin Block <bblock@linux.ibm.com>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Use a structure member to track the SCSI command submitter

Conditional statements are faster than indirect calls. Use a structure
member to track the SCSI command submitter such that later patches can call
scsi_done(scmd) instead of scmd->scsi_done(scmd).

The asymmetric behavior that scsi_send_eh_cmnd() sets the submission
context to the SCSI error handler and that it does not restore the
submission context to the SCSI core is retained.

Link: https://lore.kernel.org/r/20211007202923.2174984-2-bvanassche@acm.org
Cc: Hannes Reinecke <hare@suse.com>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Reviewed-by: Benjamin Block <bblock@linux.ibm.com>
Reviewed-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix spelling in a source code comment

The typo in this source code comment makes the comment confusing. Clear up
the confusion by fixing the typo.

Link: https://lore.kernel.org/r/20210929182318.2060489-1-bvanassche@acm.org
Fixes: bc85dc500f9d ("scsi: remove scsi_end_request")
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix scsi_mode_select() buffer length handling

The MODE SELECT(6) command allows handling mode page buffers that are up to
255 bytes, including the 4 byte header needed in front of the page
buffer. For requests larger than this limit, automatically use the MODE
SELECT(10) command.

In both cases, since scsi_mode_select() adds the mode select page header,
checks on the buffer length value must include this header size to avoid
overflows of the command CDB allocation length field.

While at it, use put_unaligned_be16() for setting the header block
descriptor length and CDB allocation length when using MODE SELECT(10).

[mkp: fix MODE SENSE vs. MODE SELECT confusion]

Link: https://lore.kernel.org/r/20210820070255.682775-3-damien.lemoal@wdc.com
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix scsi_mode_sense() buffer length handling

Several problems exist with scsi_mode_sense() buffer length handling:

1) The allocation length field of the MODE SENSE(10) command is 16-bits,
occupying bytes 7 and 8 of the CDB. With this command, access to mode
pages larger than 255 bytes is thus possible. However, the CDB
allocation length field is set by assigning len to byte 8 only, thus
truncating buffer length larger than 255.

2) If scsi_mode_sense() is called with len smaller than 8 with
sdev->use_10_for_ms set, or smaller than 4 otherwise, the buffer length
is increased to 8 and 4 respectively, and the buffer is zero filled
with these increased values, thus corrupting the memory following the
buffer.

Fix these 2 problems by using put_unaligned_be16() to set the allocation
length field of MODE SENSE(10) CDB and by returning an error when len is
too small.

Furthermore, if len is larger than 255B, always try MODE SENSE(10) first,
even if the device driver did not set sdev->use_10_for_ms. In case of
invalid opcode error for MODE SENSE(10), access to mode pages larger than
255 bytes are not retried using MODE SENSE(6). To avoid buffer length
overflows for the MODE_SENSE(10) case, check that len is smaller than 65535
bytes.

While at it, also fix the folowing:

* Use get_unaligned_be16() to retrieve the mode data length and block
descriptor length fields of the mode sense reply header instead of using
an open coded calculation.

* Fix the kdoc dbd argument explanation: the DBD bit stands for Disable
Block Descriptor, which is the opposite of what the dbd argument
description was.

Link: https://lore.kernel.org/r/20210820070255.682775-2-damien.lemoal@wdc.com
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: remove blk_{get,put}_request

These are now pointless wrappers around blk_mq_{alloc,free}_request,
so remove them.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20211025070517.1548584-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: remove QUEUE_FLAG_SCSI_PASSTHROUGH

Export scsi_device_from_queue for use with pktcdvd and use that instead
of the otherwise unused QUEUE_FLAG_SCSI_PASSTHROUGH queue flag.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20211021060607.264371-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: add a scsi_alloc_request helper

Add a new helper that calls blk_get_request and initializes the
scsi_request to avoid the indirect call through ->.initialize_rq_fn.

Note that this makes the pktcdvd driver depend on the SCSI core, but
given that only SCSI devices support SCSI passthrough requests that
is not a functional change.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20211021060607.264371-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: add a struct io_comp_batch argument to fops->iopoll()

struct io_comp_batch contains a list head and a completion handler, which
will allow completions to more effciently completed batches of IO.

For now, no functional changes in this patch, we just define the
io_comp_batch structure and add the argument to the file_operations iopoll
handler.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move integrity handling out of <linux/blkdev.h>

Split the integrity/metadata handling definitions out into a new header.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Link: https://lore.kernel.org/r/20210920123328.1399408-17-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Remove scsi_cmnd.tag

It is never read, so get rid of it.

Link: https://lore.kernel.org/r/1628862553-179450-4-git-send-email-john.garry@huawei.com
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove the request member from struct scsi_cmnd

Since all scsi_cmnd.request users are gone, remove the request pointer
from struct scsi_cmnd.

Link: https://lore.kernel.org/r/20210809230355.8186-53-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Use scsi_cmd_to_rq() instead of scsi_cmnd.request

Prepare for removal of the request pointer by using scsi_cmd_to_rq()
instead. Cast away constness where necessary when passing a SCSI command
pointer to scsi_cmd_to_rq(). This patch does not change any functionality.

Link: https://lore.kernel.org/r/20210809230355.8186-3-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: scsi_ioctl: Remove scsi_req_init()

Merge scsi_req_init() into its only caller.

Link: https://lore.kernel.org/r/20210724072033.1284840-16-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix the documentation of the scsi_execute() time parameter

The unit of the scsi_execute() timeout parameter is 1/HZ seconds instead of
one second, just like the timeouts used in the block layer. Fix the
documentation header above the definition of the scsi_execute() macro.

Link: https://lore.kernel.org/r/20210709202638.9480-3-bvanassche@acm.org
Fixes: "[SCSI] use scatter lists for all block pc requests and simplify hw handlers" # v2.6.16.28
Cc: "James E.J. Bottomley" <jejb@linux.ibm.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: John Garry <john.garry@huawei.com>
Reviewed-by: Avri Altman <avri.altman@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Retry I/O for Notify (Enable Spinup) Required error

If the device is power-cycled, it takes time for the initiator to transmit
the periodic NOTIFY (ENABLE SPINUP) SAS primitive, and for the device to
respond to the primitive to become ACTIVE. Retry the I/O request to allow
the device time to become ACTIVE.

Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20210629155826.48441-1-quat.le@oracle.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Quat Le <quat.le@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Inline scsi_mq_alloc_queue()

Since scsi_mq_alloc_queue() only has one caller, inline it. This change was
suggested by Christoph Hellwig.

Link: https://lore.kernel.org/r/20210622024654.12543-1-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Ed Tsai <ed.tsai@mediatek.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: remove REQ_OP_SCSI_{IN,OUT}

With the legacy IDE driver gone drivers now use either REQ_OP_DRV_*
or REQ_OP_SCSI_*, so unify the two concepts of passthrough requests
into a single one.

Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Drop obsolete Linux-specific SCSI status codes

Originally the SCSI subsystem has been using 'special' SCSI status codes,
which were the SAM-specified ones but shifted by 1. As most drivers have
now been modified to use the SAM-specified ones, having two nearly
identical sets of definitions only causes confusion.

The Linux-specifed SCSI status codes have been marked obsolete for several
years so drop them and use the SAM-specified status codes throughout.

Link: https://lore.kernel.org/r/20210427083046.31620-41-hare@suse.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Douglas Gilbert <dgilbert@interlog.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Kill message byte

Remove last vestiges of SCSI status message bytes.

Link: https://lore.kernel.org/r/20210427083046.31620-39-hare@suse.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Kill DRIVER_SENSE

Replace the check for DRIVER_SENSE with a check for
scsi_status_is_check_condition().

Audit all callsites to ensure the SAM status is set correctly. For
backwards compability move the DRIVER_SENSE definition to sg.h, and update
sg, bsg, and scsi_ioctl to set the DRIVER_SENSE driver_status whenever
SAM_STAT_CHECK_CONDITION is present.

[mkp: fix zeroday srp warning]

Link: https://lore.kernel.org/r/20210427083046.31620-10-hare@suse.de
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

fix

scsi: core: Introduce scsi_status_is_check_condition()

Add a helper function scsi_status_is_check_condition() to encapsulate the
frequent checks for SAM_STAT_CHECK_CONDITION.

Link: https://lore.kernel.org/r/20210427083046.31620-9-hare@suse.de
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Introduce scsi_build_sense()

Introduce scsi_build_sense() as a wrapper around scsi_build_sense_buffer()
to format the buffer and set the correct SCSI status.

Link: https://lore.kernel.org/r/20210427083046.31620-8-hare@suse.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Stop using DRIVER_ERROR

Return the actual error code in __scsi_execute() (which, according to the
documentation, should have happened anyway). And audit all callers to cope
with negative return values from __scsi_execute() and friends.

[mkp: resolve conflict and return bool]

Link: https://lore.kernel.org/r/20210427083046.31620-7-hare@suse.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Reshuffle response handling in scsi_mode_sense()

Reshuffle response handling in scsi_mode_sense() to make the code easier to
follow.

[mkp: fix build]

Link: https://lore.kernel.org/r/20210427083046.31620-5-hare@suse.de
Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fixup calling convention for scsi_mode_sense()

The description for scsi_mode_sense() claims to return the number of valid
bytes on success, which is not what the code does. Additionally there is
no gain in returning the SCSI status, as everything the callers do is to
check against scsi_result_is_good(), which is what scsi_mode_sense() does
already. So change the calling convention to return a standard error code
on failure, and 0 on success, and adapt the description and all callers.

Link: https://lore.kernel.org/r/20210427083046.31620-4-hare@suse.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Introduce enum scsi_disposition

Improve readability of the code in the SCSI core by introducing an
enumeration type for the values used internally that decide how to continue
processing a SCSI command. The eh_*_handler return values have not been
changed because that would involve modifying all SCSI drivers.

The output of the following command has been inspected to verify that no
out-of-range values are assigned to a variable of type enum
scsi_disposition:

KCFLAGS=-Wassign-enum make CC=clang W=1 drivers/scsi/

Link: https://lore.kernel.org/r/20210415220826.29438-6-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename scsi_softirq_done() into scsi_complete()

Commit 320ae51feed5 ("blk-mq: new multi-queue block IO queueing mechanism";
v3.13) introduced a code path that calls the blk-mq completion function
from interrupt context. scsi-mq was introduced by commit d285203cf647
("scsi: add support for a blk-mq based I/O path."; v3.17).

Since the introduction of scsi-mq, scsi_softirq_done() can be called from
interrupt context. That made the name of the function misleading, rename it
to scsi_complete().

Link: https://lore.kernel.org/r/20210415220826.29438-4-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Make the scsi_alloc_sgtables() documentation more accurate

The current scsi_alloc_sgtables() documentation does not accurately explain
what this function does. Hence improve the documentation of this function.

Link: https://lore.kernel.org/r/20210415220826.29438-2-bvanassche@acm.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: remove the unchecked_isa_dma flag

Remove the unchecked_isa_dma now that all users are gone.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20210331073001.46776-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Set shost as hctx driver_data

hctx->driver_data is not set for SCSI currently. Set hctx->driver_data =
shost.

Link: https://lore.kernel.org/r/20210215074048.19424-6-kashyap.desai@broadcom.com
Suggested-by: John Garry <john.garry@huawei.com>
Reviewed-by: John Garry <john.garry@huawei.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add mq_poll support to SCSI layer

Currently IOPOLL support is only available in block layer. This patch
adds mq_poll support to the SCSI layer.

Link: https://lore.kernel.org/r/20210215074048.19424-2-kashyap.desai@broadcom.com
Cc: sumit.saxena@broadcom.com
Cc: chandrakanth.patil@broadcom.com
Cc: linux-block@vger.kernel.org
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Replace sdev->device_busy with sbitmap

SCSI currently uses an atomic variable to track queue depth for each
attached device. The queue depth depends on many factors such as transport
type and device implementation. In addition, the SCSI device queue depth is
not a static entity but changes over time as a result of congestion
management.

While blk-mq currently tracks queue depth for each hctx, it can't easily be
changed to accommodate the SCSI per-device requirement.

The current approach of using an atomic variable doesn't scale well when
there are lots of CPU cores and the disk is very fast. IOPS can be
substantially impacted by the atomic in the hot path.

Replace the atomic variable sdev->device_busy with an sbitmap for tracking
the SCSI device queue depth.

It has been observed that IOPS is improved ~30% by this patchset in the
following test:

1) test machine(32 logical CPU cores)
Thread(s) per core: 2
Core(s) per socket: 8
Socket(s): 2
NUMA node(s): 2
Model name: Intel(R) Xeon(R) Silver 4110 CPU @ 2.10GHz

2) setup scsi_debug:
modprobe scsi_debug virtual_gb=128 max_luns=1 submit_queues=32 delay=0 max_queue=256

3) fio script:
fio --rw=randread --size=128G --direct=1 --ioengine=libaio --iodepth=2048 \
--numjobs=32 --bs=4k --group_reporting=1 --group_reporting=1 --runtime=60 \
--loops=10000 --name=job1 --filename=/dev/sdN

[mkp: fix device_busy reference in mpt3sas]

Link: https://lore.kernel.org/r/20210122023317.687987-14-ming.lei@redhat.com
Link: https://lore.kernel.org/linux-block/20200119071432.18558-6-ming.lei@redhat.com/
Cc: Omar Sandoval <osandov@fb.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Tested-by: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add scsi_device_busy() wrapper

Add scsi_device_busy() helper to prepare drivers for tracking device queue
depth via sbitmap_queue.

Link: https://lore.kernel.org/r/20210122023317.687987-12-ming.lei@redhat.com
Cc: Omar Sandoval <osandov@fb.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Tested-by: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: blk-mq: Return budget token from .get_budget callback

SCSI uses a global atomic variable to track queue depth for each
LUN/request queue.

This doesn't scale well when there are lots of CPU cores and the disk is
very fast. It has been observed that IOPS is affected a lot by tracking
queue depth via sdev->device_busy in the I/O path.

Return budget token from .get_budget callback. The budget token can be
passed to driver so that we can replace the atomic variable with
sbitmap_queue and alleviate the scaling problems that way.

Link: https://lore.kernel.org/r/20210122023317.687987-9-ming.lei@redhat.com
Cc: Omar Sandoval <osandov@fb.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Tested-by: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: blk-mq: Add callbacks for storing & retrieving budget token

Since SCSI is the only driver which requires dispatch budget move the token
from struct request to struct scsi_cmnd.

Link: https://lore.kernel.org/r/20210122023317.687987-8-ming.lei@redhat.com
Cc: Omar Sandoval <osandov@fb.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Tested-by: Sumanesh Samanta <sumanesh.samanta@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Add a new error code DID_TRANSPORT_MARGINAL in scsi.h

Add code in scsi_result_to_blk_status to translate a new error
DID_TRANSPORT_MARGINAL to the corresponding blk_status_t i.e
BLK_STS_TRANSPORT.

Add DID_TRANSPORT_MARGINAL case to scsi_decide_disposition().

Link: https://lore.kernel.org/r/1609969748-17684-2-git-send-email-muneendra.kumar@broadcom.com
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Muneendra Kumar <muneendra.kumar@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: remove unnecessary argument from blk_execute_rq

We can remove 'q' from blk_execute_rq as well after the previous change
in blk_execute_rq_nowait.

And more importantly it never really was needed to start with given
that we can trivial derive it from struct request.

Cc: linux-scsi@vger.kernel.org
Cc: virtualization@lists.linux-foundation.org
Cc: linux-ide@vger.kernel.org
Cc: linux-mmc@vger.kernel.org
Cc: linux-nvme@lists.infradead.org
Cc: linux-nfs@vger.kernel.org
Acked-by: Ulf Hansson <ulf.hansson@linaro.org> # for mmc
Signed-off-by: Guoqing Jiang <guoqing.jiang@cloud.ionos.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Only process PM requests if rpm_status != RPM_ACTIVE

Instead of submitting all SCSI commands submitted with scsi_execute() to a
SCSI device if rpm_status != RPM_ACTIVE, only submit RQF_PM (power
management requests) if rpm_status != RPM_ACTIVE. This patch makes the SCSI
core handle the runtime power management status (rpm_status) as it should
be handled.

Link: https://lore.kernel.org/r/20201209052951.16136-7-bvanassche@acm.org
Cc: Can Guo <cang@codeaurora.org>
Cc: Stanley Chu <stanley.chu@mediatek.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Martin Kepplinger <martin.kepplinger@puri.sm>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Can Guo <cang@codeaurora.org>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix race between handling STS_RESOURCE and completion

When queuing I/O request to LLD, STS_RESOURCE may be returned because:

- Host is in recovery or blocked

- Target queue throttling or target is blocked

- LLD rejection

In these scenarios BLK_STS_DEV_RESOURCE is returned to the block layer to
avoid an unnecessary re-run of the queue. However, all of the requests
queued to this SCSI device may complete immediately after reading
'sdev->device_busy' and BLK_STS_DEV_RESOURCE is returned to block layer. In
that case the current I/O won't get a chance to get queued since it is
invisible at that time for both scsi_run_queue_async() and blk-mq's
RESTART.

Fix the issue by not returning BLK_STS_DEV_RESOURCE in this situation.

Link: https://lore.kernel.org/r/20201202100419.525144-1-ming.lei@redhat.com
Fixes: 86ff7c2a80cd ("blk-mq: introduce BLK_STS_DEV_RESOURCE")
Cc: Hannes Reinecke <hare@suse.com>
Cc: Sumit Saxena <sumit.saxena@broadcom.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan Milne <emilne@redhat.com>
Cc: Long Li <longli@microsoft.com>
Reported-by: John Garry <john.garry@huawei.com>
Tested-by: "chenxiang (M)" <chenxiang66@hisilicon.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Replace while-loop by for-loop in scsi_vpd_lun_id()

This makes the code slightly more readable.

Link: https://lore.kernel.org/r/20201029170846.14786-2-mwilck@suse.com
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix VPD LUN ID designator priorities

The current implementation of scsi_vpd_lun_id() uses the designator length
as an implicit measure of priority. This works most of the time, but not
always. For example, some Hitachi storage arrays return this in VPD 0x83:

VPD INQUIRY: Device Identification page
Designation descriptor number 1, descriptor length: 24
designator_type: T10 vendor identification, code_set: ASCII
associated with the Addressed logical unit
vendor id: HITACHI
vendor specific: 5030C3502025
Designation descriptor number 2, descriptor length: 6
designator_type: vendor specific [0x0], code_set: Binary
associated with the Target port
vendor specific: 08 03
Designation descriptor number 3, descriptor length: 20
designator_type: NAA, code_set: Binary
associated with the Addressed logical unit
NAA 6, IEEE Company_id: 0x60e8
Vendor Specific Identifier: 0x7c35000
Vendor Specific Identifier Extension: 0x30c35000002025
[0x60060e8007c350000030c35000002025]

The current code would use the first descriptor because it's longer than
the NAA descriptor. But this is wrong, the kernel is supposed to prefer NAA
descriptors over T10 vendor ID. Designator length should only be used to
compare designators of the same type.

This patch addresses the issue by separating designator priority and
length.

Link: https://lore.kernel.org/r/20201029170846.14786-1-mwilck@suse.com
Fixes: 9983bed3907c ("scsi: Add scsi_vpd_lun_id()")
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Return BLK_STS_AGAIN for ALUA transitioning

Whenever we encounter a sense code of ALUA transitioning in
scsi_io_completion() it means that the SCSI midlayer ran out of retries
trying to wait for ALUA transitioning. In these cases we should be passing
up the error, but signalling that the I/O might be retried, preferably on
another path. So return BLK_STS_AGAIN in these cases.

[mkp: typo + fallthrough]

Link: https://lore.kernel.org/r/20200930080256.90964-5-hare@suse.de
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: scsi_dh_alua: Return BLK_STS_AGAIN for ALUA transitioning state

When the ALUA state indicates transitioning we should not retry the command
immediately, but rather complete the command with BLK_STS_AGAIN to signal
the completion handler that it might be retried. This allows multipathing
to redirect the command to another path if possible, and avoid stalls
during lengthy transitioning times.

Link: https://lore.kernel.org/r/20200930080256.90964-3-hare@suse.de
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: doc: Fix some kernel-doc markups

Some identifiers have different names between their prototypes
and the kernel-doc markup.

[mkp: fix whitespace]

Link: https://lore.kernel.org/r/8ed7f149f25a363eea76e514c253c4e337c59379.1603469755.git.mchehab+huawei@kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: handle zone resources errors

ZBC or ZAC disks that have a limit on the number of open zones may fail
a zone open command or a write to a zone that is not already implicitly
or explicitly open if the total number of open zones is already at the
maximum allowed.

For these operations, instead of returning the generic BLK_STS_IOERR,
return BLK_STS_ZONE_OPEN_RESOURCE which is returned as -ETOOMANYREFS to
the I/O issuer, allowing the device user to act appropriately on these
relatively benign zone resource errors.

Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Set sc_data_direction to DMA_NONE for no-transfer commands

No having the special DMA_NONE logic makes libata rather unhappy.

Link: https://lore.kernel.org/r/20201008200611.1818099-3-hch@lst.de
Fixes: 40b93836a136 ("scsi: core: Use rq_dma_dir in scsi_setup_cmnd()")
Reported-by: Qian Cai <cai@redhat.com>
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Only start the request just before dispatching

This has no change in behavior, but improves the accounting a bit.

Link: https://lore.kernel.org/r/20201005084130.143273-11-hch@lst.de
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove scsi_setup_cmnd() and scsi_setup_fs_cmnd()

Move this trivial functionality into scsi_prepare_cmd() instead of
splitting it over multiple small functions, and update the comments to
better document passthrough commands as the special case.

Link: https://lore.kernel.org/r/20201005084130.143273-10-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Clean up allocation and freeing of sgtables

Rename scsi_init_io() to scsi_alloc_sgtables(), and ensure callers call
scsi_free_sgtables() to cleanup failures close to scsi_init_io() instead of
leaking it down the generic I/O submission path.

Link: https://lore.kernel.org/r/20201005084130.143273-9-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename scsi_mq_prep_fn() to scsi_prepare_cmd()

The old name is rather confusing now that the the legacy prep_fn is gone.

Link: https://lore.kernel.org/r/20201005084130.143273-8-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename scsi_prep_state_check() to scsi_device_state_check()

The old name is rather confusing now that the the legacy prep_fn is gone.

Link: https://lore.kernel.org/r/20201005084130.143273-7-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Use rq_dma_dir in scsi_setup_cmnd()

Link: https://lore.kernel.org/r/20201005084130.143273-6-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Move command size detection out of the fast path

We only need to detect the command size for ioctl request from userspace,
which is limited to the passthrough path. Move the check there instead of
doing it for all queuecommand invocations.

Link: https://lore.kernel.org/r/20201005084130.143273-4-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove scsi_init_cmd_errh

There is no good reason to keep this functionality as a separate function,
just merge it into the only caller.

Link: https://lore.kernel.org/r/20201005084130.143273-3-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Don't export scsi_device_from_queue()

This function is only used by code built into scsi_mod.ko.

Link: https://lore.kernel.org/r/20201005084130.143273-2-hch@lst.de
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Add host and host template flag 'host_tagset'

Add Host and host template flag 'host_tagset' so hostwide tagset can be
shared on multiple reply queues after the SCSI device's reply queue is
converted to blk-mq hw queue.

[jpg: Update comment on .can_queue and add Scsi_Host.host_tagset]

Signed-off-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: John Garry <john.garry@huawei.com>
Tested-by: Don Brace<don.brace@microsemi.com> #SCSI resv cmds patches used
Tested-by: Douglas Gilbert <dgilbert@interlog.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Add limitless cmd retry support

Add infinite retry support to SCSI midlayer by combining common checks for
retries into some helper functions, and then checking for the
-1/SCSI_CMD_RETRIES_NO_LIMIT.

Link: https://lore.kernel.org/r/1601566554-26752-2-git-send-email-michael.christie@oracle.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Only re-run queue in scsi_end_request() if device queue is busy

The request queue is currently run unconditionally in scsi_end_request() if
both target queue and host queue are ready.

Recently Long Li reported that cost of a queue run can be very heavy in
case of high queue depth. Improve this situation by only running the
request queue when this LUN is busy.

Link: https://lore.kernel.org/r/20200910075056.36509-1-ming.lei@redhat.com
Reported-by: Long Li <longli@microsoft.com>
Tested-by: Long Li <longli@microsoft.com>
Tested-by: Kashyap Desai <kashyap.desai@broadcom.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

treewide: Use fallthrough pseudo-keyword

Replace the existing /* fall through */ comments and its variants with
the new pseudo-keyword macro fallthrough[1]. Also, remove unnecessary
fall-through markings when it is the case.

[1] https://www.kernel.org/doc/html/v5.7/process/deprecated.html?highlight=fallthrough#implicit-switch-case-fall-through

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>

scsi: core: Add missing scsi_device_put() in scsi_host_block()

The scsi_host_block() case was missing in commit 4dea170f4fb2 ("scsi: core:
Fix incorrect usage of shost_for_each_device").

Link: https://lore.kernel.org/r/20200717090921.29243-1-yebin10@huawei.com
Fixes: 2bb955840c1d ("scsi: core: add scsi_host_(block,unblock) helper function")
Fixes: 4dea170f4fb2 ("scsi: core: Fix incorrect usage of shost_for_each_device")
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Run queue in case of I/O resource contention failure

I/O requests may be held in scheduler queue because of resource contention.
The starvation scenario was handled properly in the regular completion
path but we failed to account for it during I/O submission. This lead to
the hang captured below. Make sure we run the queue when resource
contention is encountered in the submission path.

[ 39.054963] scsi 13:0:0:0: rejecting I/O to dead device
[ 39.058700] scsi 13:0:0:0: rejecting I/O to dead device
[ 39.087855] sd 13:0:0:1: [sdd] Synchronizing SCSI cache
[ 39.088909] scsi 13:0:0:1: rejecting I/O to dead device
[ 39.095351] scsi 13:0:0:1: rejecting I/O to dead device
[ 39.096962] scsi 13:0:0:1: rejecting I/O to dead device
[ 247.021859] INFO: task scsi-stress-rem:813 blocked for more than 122 seconds.
[ 247.023258] Not tainted 5.8.0-rc2 #8
[ 247.024069] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 247.025331] scsi-stress-rem D 0 813 802 0x00004000
[ 247.025334] Call Trace:
[ 247.025354] __schedule+0x504/0x55f
[ 247.027987] schedule+0x72/0xa8
[ 247.027991] blk_mq_freeze_queue_wait+0x63/0x8c
[ 247.027994] ? do_wait_intr_irq+0x7a/0x7a
[ 247.027996] blk_cleanup_queue+0x4b/0xc9
[ 247.028000] __scsi_remove_device+0xf6/0x14e
[ 247.028002] scsi_remove_device+0x21/0x2b
[ 247.029037] sdev_store_delete+0x58/0x7c
[ 247.029041] kernfs_fop_write+0x10d/0x14f
[ 247.031281] vfs_write+0xa2/0xdf
[ 247.032670] ksys_write+0x6b/0xb3
[ 247.032673] do_syscall_64+0x56/0x82
[ 247.034053] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 247.034059] RIP: 0033:0x7f69f39e9008
[ 247.036330] Code: Bad RIP value.
[ 247.036331] RSP: 002b:00007ffdd8116498 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[ 247.037613] RAX: ffffffffffffffda RBX: 0000000000000002 RCX: 00007f69f39e9008
[ 247.039714] RDX: 0000000000000002 RSI: 000055cde92a0ab0 RDI: 0000000000000001
[ 247.039715] RBP: 000055cde92a0ab0 R08: 000000000000000a R09: 00007f69f3a79e80
[ 247.039716] R10: 000000000000000a R11: 0000000000000246 R12: 00007f69f3abb780
[ 247.039717] R13: 0000000000000002 R14: 00007f69f3ab6740 R15: 0000000000000002

Link: https://lore.kernel.org/r/20200720025435.812030-1-ming.lei@redhat.com
Cc: linux-block@vger.kernel.org
Cc: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

blk-mq: pass request queue into get/put budget callback

blk-mq budget is abstract from scsi's device queue depth, and it is
always per-request-queue instead of hctx.

It can be quite absurd to get a budget from one hctx, then dequeue a
request from scheduler queue, and this request may not belong to this
hctx, at least for bfq and deadline.

So fix the mess and always pass request queue to get/put budget
callback.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Baolin Wang <baolin.wang7@gmail.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Cc: Sagi Grimberg <sagi@grimberg.me>
Cc: Baolin Wang <baolin.wang7@gmail.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: move failure injection out of blk_mq_complete_request

Move the call to blk_should_fake_timeout out of blk_mq_complete_request
and into the drivers, skipping call sites that are obvious error
handlers, and remove the now superflous blk_mq_force_complete_rq helper.
This ensures we don't keep injecting errors into completions that just
terminate the Linux request after the hardware has been reset or the
command has been aborted.

Reviewed-by: Daniel Wagner <dwagner@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Fix formatting errors in scsi_lib.c

Delete trailing whitespace, multiple blank lines, and make switch/case be
at the same indentation.

Link: https://lore.kernel.org/r/20200619154117.10262-3-huobean@gmail.com
Signed-off-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove scsi_sdb_cache

After commit f664a3cc17b7 ("scsi: kill off the legacy IO path"),
scsi_sdb_cache is not used anymore. Remove it.

Link: https://lore.kernel.org/r/20200619154117.10262-2-huobean@gmail.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Bean Huo <beanhuo@micron.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Refactor scsi_mq_setup_tags function

shost->tag_set is used too many times, introduce temporary parameter
tag_set instead of &shost->tag_set.

Link: https://lore.kernel.org/r/20200518074732.39679-1-yebin10@huawei.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix incorrect usage of shost_for_each_device

shost_for_each_device(sdev, shost) \
for ((sdev) = __scsi_iterate_devices((shost), NULL); \
(sdev); \
(sdev) = __scsi_iterate_devices((shost), (sdev)))

When terminating shost_for_each_device() iteration with break or return,
scsi_device_put() should be used to prevent stale scsi device references
from being left behind.

Link: https://lore.kernel.org/r/20200518074420.39275-1-yebin10@huawei.com
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Ye Bin <yebin10@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Introduce REQ_OP_ZONE_APPEND

Define REQ_OP_ZONE_APPEND to append-write sectors to a zone of a zoned
block device. This is a no-merge write operation.

A zone append write BIO must:
* Target a zoned block device
* Have a sector position indicating the start sector of the target zone
* The target zone must be a sequential write zone
* The BIO must not cross a zone boundary
* The BIO size must not be split to ensure that a single range of LBAs
is written with a single command.

Implement these checks in generic_make_request_checks() using the
helper function blk_check_zone_append(). To avoid write append BIO
splitting, introduce the new max_zone_append_sectors queue limit
attribute and ensure that a BIO size is always lower than this limit.
Export this new limit through sysfs and check these limits in bio_full().

Also when a LLDD can't dispatch a request to a specific zone, it
will return BLK_STS_ZONE_RESOURCE indicating this request needs to
be delayed, e.g. because the zone it will be dispatched to is still
write-locked. If this happens set the request aside in a local list
to continue trying dispatching requests such as READ requests or a
WRITE/ZONE_APPEND requests targetting other zones. This way we can
still keep a high queue depth without starving other requests even if
one request can't be served due to zone write-locking.

Finally, make sure that the bio sector position indicates the actual
write position as indicated by the device on completion.

Signed-off-by: Keith Busch <kbusch@kernel.org>
[ jth: added zone-append specific add_page and merge_page helpers ]
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: free sgtables in case command setup fails

In case scsi_setup_fs_cmnd() fails we're not freeing the sgtables allocated
by scsi_init_io(), thus we leak the allocated memory.

Free the sgtables allocated by scsi_init_io() in case scsi_setup_fs_cmnd()
fails.

Technically scsi_setup_scsi_cmnd() does not suffer from this problem as it
can only fail if scsi_init_io() fails, so it does not have sgtables
allocated. But to maintain symmetry and as a measure of defensive
programming, free the sgtables on scsi_setup_scsi_cmnd() failure as well.
scsi_mq_free_sgtables() has safeguards against double-freeing of memory so
this is safe to do.

While we're at it, rename scsi_mq_free_sgtables() to scsi_free_sgtables().

Link: https://bugzilla.kernel.org/show_bug.cgi?id=205595
Link: https://lore.kernel.org/r/20200428104605.8143-2-johannes.thumshirn@wdc.com
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Daniel Wagner <dwagner@suse.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: doc: Change function comments to kernel-doc style

Despite of functions being documented, they are not in the kernel-doc
specification, and could not be included in kernel documentation. Change
the style of functions comments to be compliant to the kernel-doc style.
When the function comments are outdated, update then.

[mkp: a few edits]

Link: https://lore.kernel.org/r/20200419050148.33371-1-andrealmeid@collabora.com
Signed-off-by: André Almeida <andrealmeid@collabora.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Avoid calling synchronize_rcu() for each device in scsi_host_block()

scsi_host_block() calls scsi_internal_device_block() for each scsi_device and
scsi_internal_device_block() calls blk_mq_quiesce_queue() for each LUN.

Since synchronize_rcu() is called from blk_mq_quiesce_queue(), this can cause
substantial slowdowns on systems with many LUNs.

Use scsi_internal_device_block_nowait() to implement scsi_host_block() so it
is sufficient to run synchronize_rcu() once. This is safe since SCSI does not
set the BLK_MQ_F_BLOCKING flag.

[mkp: commit desc and comment tweaks]

Link: https://lore.kernel.org/r/20200423020713.332743-1-ming.lei@redhat.com
Cc: Steffen Maier <maier@linux.ibm.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dexuan Cui <decui@microsoft.com>
Cc: Hannes Reinecke <hare@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: move dma_pad handling from blk_rq_map_sg into the callers

There are only two callers of blk_rq_map_sg/__blk_rq_map_sg that set
the dma_pad value in the queue. Move the handling into those callers
instead of burdening the common code, and move the ->extra_len field
from struct request to struct scsi_cmnd.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: move dma drain handling to scsi

Don't burden the common block code with with specifics of the libata DMA
draining mechanism. Instead move most of the code to the scsi midlayer.

That also means the nr_phys_segments adjustments in the blk-mq fast path
can go away entirely, given that SCSI never looks at nr_phys_segments
after mapping the request to a scatterlist.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: merge scsi_init_sgtable into scsi_init_io

scsi_init_io is the only caller of scsi_init_sgtable. Merge the two
function to make upcoming changes a little easier.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Allow the state change from SDEV_QUIESCE to SDEV_BLOCK

The APIs scsi_host_block()/scsi_host_unblock() were recently added by
commit 2bb955840c1d ("scsi: core: add scsi_host_(block,unblock) helper
function") and so far the APIs are only used by: commit 3d3ca53b1639
("scsi: aacraid: use scsi_host_(block,unblock) to block I/O").

However, from reading the code, I think the APIs don't really work for
aacraid, because, in the resume path of hibernation, when aac_suspend() ->
scsi_host_block() is called, scsi_device_quiesce() has set the state to
SDEV_QUIESCE, so aac_suspend() -> scsi_host_block() returns -EINVAL.

Fix the issue by allowing the state change.

Link: https://lore.kernel.org/r/1587170445-50013-1-git-send-email-decui@microsoft.com
Fixes: 2bb955840c1d ("scsi: core: add scsi_host_(block,unblock) helper function")
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

Revert "scsi: core: run queue if SCSI device queue isn't ready and queue is idle"

This reverts commit 7e70aa789d4a0c89dbfbd2c8a974a4df717475ec.

Now that we have the patches ("blk-mq: In blk_mq_dispatch_rq_list()
"no budget" is a reason to kick") and ("blk-mq: Rerun dispatching in
the case of budget contention") we should no longer need the fix in
the SCSI code. Revert it, resolving conflicts with other patches that
have touched this code.

With this revert (and the two new patches) I can run the script that
was in commit 7e70aa789d4a ("scsi: core: run queue if SCSI device
queue isn't ready and queue is idle") in a loop with no failure. If I
do this revert without the two new patches I can easily get a failure.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: avoid repetitive logging of device offline messages

Large queues of I/O to offline devices that are eventually submitted when
devices are unblocked result in a many repeated "rejecting I/O to offline
device" messages. These messages can fill up the dmesg buffer in crash
dumps so no useful prior messages remain. In addition, if a serial console
is used, the flood of messages can cause a hard lockup in the console code.

Introduce a flag indicating the message has already been logged for the
device, and reset the flag when scsi_device_set_state() changes the device
state.

Link: https://lore.kernel.org/r/20200311143930.20674-1-emilne@redhat.com
Reviewed-by: Bart van Assche <bvanassche@acm.org>
Signed-off-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Introduce {init,exit}_cmd_priv()

The current behavior of the SCSI core is to clear driver-private data
before preparing a request for submission to the SCSI LLD. Make it possible
for SCSI LLDs to disable clearing of driver-private data.

These hooks will be used by a later patch, namely "scsi: ufs: Let the SCSI
core allocate per-command UFS data".

Link: https://lore.kernel.org/r/20200123035637.21848-2-bvanassche@acm.org
Cc: Tomas Winkler <tomas.winkler@intel.com>
Cc: Stanley Chu <stanley.chu@mediatek.com>
Cc: Bean Huo <beanhuo@micron.com>
Cc: Avri Altman <avri.altman@wdc.com>
Cc: Can Guo <cang@codeaurora.org>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Johannes Thumshirn <jth@kernel.org>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove cmd_list functionality

Remove cmd_list functionality; no users left. With that the
scsi_put_command() becomes empty, so remove that one, too.

Link: https://lore.kernel.org/r/20200228075318.91255-14-hare@suse.de
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart van Assche <bvanassche@acm.org>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add scsi_host_(block,unblock) helper function

Add helper functions to call scsi_internal_device_block()/
scsi_internal_device_unblock() for all attached devices on a SCSI host.

Link: https://lore.kernel.org/r/20200228075318.91255-9-hare@suse.de
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Adjust DBD setting in MODE SENSE for caching mode page per LLD

UFS JEDEC standards require DBD field to be set to 1 in mode sense
command. This patch allows LLD to define the setting of DBD, if
required.

Link: https://lore.kernel.org/r/0101016ed3d643f9-ffd45d6c-c593-4a13-a18f-a32da3d3bb97-000000@us-west-2.amazonses.com
Reviewed-by: Avri Altman <avri.altman@wdc.com>
Signed-off-by: Can Guo <cang@codeaurora.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Handle drivers which set sg_tablesize to zero

In scsi_mq_setup_tags(), cmd_size is calculated based on zero size for the
scatter-gather list in case the low level driver uses SG_NONE in its host
template.

cmd_size is passed on to the block layer for calculation of the request
size, and we've seen NULL pointer dereference errors from the block layer
in drivers where SG_NONE is used and a mq IO scheduler is active,
apparently as a consequence of this (see commit 68ab2d76e4be ("scsi:
cxlflash: Set sg_tablesize to 1 instead of SG_NONE"), and a recent patch by
Finn Thain converting the three m68k NFR5380 drivers to avoid setting
SG_NONE).

Try to avoid these errors by accounting for at least one sg list entry when
calculating cmd_size, regardless of whether the low level driver set a zero
sg_tablesize.

Tested on 030 m68k with the atari_scsi driver - setting sg_tablesize to
SG_NONE no longer results in a crash when loading this driver.

CC: Finn Thain <fthain@telegraphics.com.au>
Link: https://lore.kernel.org/r/1572922150-4358-1-git-send-email-schmitzmic@gmail.com
Signed-off-by: Michael Schmitz <schmitzmic@gmail.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: avoid host-wide host_busy counter for scsi_mq

It isn't necessary to check the host depth in scsi_queue_rq() any more
since it has been respected by blk-mq before calling scsi_queue_rq() via
getting driver tag.

Lots of LUNs may attach to same host and per-host IOPS may reach millions,
so we should avoid expensive atomic operations on the host-wide counter in
the IO path.

This patch implements scsi_host_busy() via blk_mq_tagset_busy_iter() with
one scsi command state for reading the count of busy IOs for scsi_mq.

It is observed that IOPS is increased by 15% in IO test on scsi_debug (32
LUNs, 32 submit queues, 1024 can_queue, libaio/dio) in a dual-socket
system.

Cc: Jens Axboe <axboe@kernel.dk>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20191025065855.6309-1-ming.lei@redhat.com
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: fix dh and multipathing for SCSI hosts without request batching

This was missing from scsi_device_from_queue() due to the introduction of
another new scsi_mq_ops_no_commit of linux-next commit 8930a6c20791 ("scsi:
core: add support for request batching") from Martin's scsi/5.4/scsi-queue
or James' scsi/misc.

Only devicehandler code seems to call scsi_device_from_queue():
*** drivers/scsi/scsi_dh.c:
scsi_dh_activate[255] sdev = scsi_device_from_queue(q);
scsi_dh_set_params[302] sdev = scsi_device_from_queue(q);
scsi_dh_attach[325] sdev = scsi_device_from_queue(q);
scsi_dh_attached_handler_name[363] sdev = scsi_device_from_queue(q);

Fixes multipath tools follow-on errors:

$ multipath -v6
...
libdevmapper: ioctl/libdm-iface.c(1887): device-mapper: reload ioctl on mpatha failed: No such device
...
mpatha: failed to load map, error 19
...

showing also as kernel messages:

device-mapper: table: 252:0: multipath: error attaching hardware handler
device-mapper: ioctl: error adding target to table

Signed-off-by: Steffen Maier <maier@linux.ibm.com>
Fixes: 8930a6c20791 ("scsi: core: add support for request batching")
Cc: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: fix missing .cleanup_rq for SCSI hosts without request batching

This was missing from scsi_mq_ops_no_commit of linux-next commit
8930a6c20791 ("scsi: core: add support for request batching") from Martin's
scsi/5.4/scsi-queue or James' scsi/misc.

See also linux-next commit b7e9e1fb7a92 ("scsi: implement .cleanup_rq
callback") from block/for-next.

Signed-off-by: Steffen Maier <maier@linux.ibm.com>
Fixes: 8930a6c20791 ("scsi: core: add support for request batching")
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Complain if scsi_target_block() fails

If scsi_target_block() fails that can break the code that calls this
function. Hence complain loudly if scsi_target_block() fails.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Make scsi_internal_device_unblock_nowait() reject invalid new_state

The only 'new_state' values passed by upstream kernel code to
scsi_internal_device_unblock_nowait() are SDEV_RUNNING and
SDEV_TRANSPORT_OFFLINE. These are the only values that should be passed to
this function. Hence check the value of the 'new_state' argument to avoid
that scsi_internal_device_unblock_nowait() would be used to trigger an
illegal SCSI device state transition. In this context 'illegal' means not
allowed by scsi_device_set_state().

Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: implement .cleanup_rq callback

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 <emilne@redhat.com>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: dm-devel@redhat.com
Cc: <stable@vger.kernel.org>
Fixes: 396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via blk_insert_cloned_request feedback")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: fix the dma_max_mapping_size call

We should only call dma_max_mapping_size for devices that have a DMA mask
set, otherwise we can run into a NULL pointer dereference that will crash
the system.

Also we need to do right shift to get the sectors from the size in bytes,
not a left shift.

Fixes: bdd17bdef7d8 ("scsi: core: take the DMA max mapping size into account")
Reported-by: Bart Van Assche <bvanassche@acm.org>
Reported-by: Ming Lei <tom.leiming@gmail.com>
Tested-by: Guilherme G. Piccoli <gpiccoli@canonical.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add support for request batching

This allows a list of requests to be issued, with the LLD only writing the
hardware doorbell when necessary, after the last request was prepared.
This is more efficient if we have lists of requests to issue, particularly
on virtualized hardware, where writing the doorbell is more expensive than
on real hardware.

The use case for this is plugged IO, where blk-mq flushes a batch of
requests all at once.

The API is the same as for blk-mq, just with blk-mq concepts tweaked to
fit the SCSI subsystem API: the "last" flag in blk_mq_queue_data becomes a
flag in scsi_cmnd, while the queue_num in the commit_rqs callback is
extracted from the hctx and passed as a parameter.

The only complication is that blk-mq uses different plugging heuristics
depending on whether commit_rqs is present or not. So we have two
different sets of blk_mq_ops and pick one depending on whether the
scsi_host template uses commit_rqs or not.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: take the DMA max mapping size into account

We need to limit the device's max_sectors to what the DMA mapping
implementation can support. If not, we risk running out of swiotlb
buffers easily.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add a host / host template field for the virt boundary

This allows drivers setting it up easily instead of branching out to block
layer calls in slave_alloc, and ensures the upgraded max_segment_size
setting gets picked up by the DMA layer.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Kashyap Desai < kashyap.desai@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Fix race on creating sense cache

When scsi_init_sense_cache(host) is called concurrently from different
hosts, each code path may find that no cache has been created and
allocate a new one. The lack of locking can lead to potentially
overriding a cache allocated by a different host.

Fix the issue by moving 'mutex_lock(&scsi_sense_cache_mutex)' before
scsi_select_sense_cache().

Fixes: 0a6ac4ee7c21 ("scsi: respect unchecked_isa_dma for blk-mq")
Cc: Stable <stable@vger.kernel.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: use scmd_printk() to print which command timed out

With a possibly faulty disk the following messages may appear in the logs:

kernel: sd 0:0:9:0: timing out command, waited 180s
kernel: sd 0:0:9:0: timing out command, waited 20s
kernel: sd 0:0:9:0: timing out command, waited 20s
kernel: sd 0:0:9:0: timing out command, waited 60s
kernel: sd 0:0:9:0: timing out command, waited 20s

This is not very informative because it's not possible to identify the
command that timed out.

This patch replaces sdev_printk() with scmd_printk().

Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Avoid that .queuecommand() gets called for a blocked SCSI device

Several SCSI transport and LLD drivers surround code that does not
tolerate concurrent calls of .queuecommand() with scsi_target_block() /
scsi_target_unblock(). These last two functions use
blk_mq_quiesce_queue() / blk_mq_unquiesce_queue() for scsi-mq request
queues to prevent concurrent .queuecommand() calls. However, that is
not sufficient to prevent .queuecommand() calls from scsi_send_eh_cmnd().
Hence surround the .queuecommand() call from the SCSI error handler with
code that avoids that .queuecommand() gets called in the blocked state.

Note: converting the .queuecommand() call in scsi_send_eh_cmnd() into
code that calls blk_get_request() + blk_execute_rq() is not an option
since scsi_send_eh_cmnd() must be able to make forward progress even
if all requests have been allocated.

Cc: Ming Lei <ming.lei@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: don't preallocate small SGL in case of NO_SG_CHAIN

The preallocated small SGL depends on SG_CHAIN so if the ARCH doesn't
support SG_CHAIN, preallocation of small SGL can't work at all.

Fix this issue by not using small preallocation in case of NO_SG_CHAIN.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: avoid preallocating big SGL for data

scsi_mq_setup_tags() preallocates a big buffer for the IO SGL. The size is
based on scsi_mq_sgl_size() which is determined based on
shost->sg_tablesize and SG_CHUNK_SIZE.

Modern DMA engines are often capable of dealing with very big segments so
the resulting scsi_mq_sgl_size() is often too big. SG_CHUNK_SIZE results in
a static 4KB SGL allocation per command.

If an HBA has lots of deep queues, preallocation for the sg list can
consume substantial amounts of memory. For lpfc, nr_hw_queues can be 70
and each queue's depth 3781. This means the resulting preallocation for
the data SGL is 70*3781*2K = 517MB.

Switch to runtime allocation for SGL for lists longer than 2 entries. This
is the approach used by NVMe PCI so it should be reasonable for SCSI as
well. Runtime SGL allocation has always been the case for the legacy I/O
path so this is nothing new.

[mkp: attempted to clarify commit desc]

Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: avoid preallocating big SGL for protection information

scsi_mq_setup_tags() currently preallocates a big buffer for protection
SGL entries. scsi_mq_sgl_size() is used to determine the size for both data
and protection information scatterlists but the protection buffer is
usually much smaller. For example, one 512-byte sector needs 8 bytes of
protection information. Given that the maximum number of sectors for one
request is 2560 (BLK_DEF_MAX_SECTORS) sectors, the max protection
information buffer size is just 20K.

The protection information segment count generally matches the number of
bios in the request. As a result, the typical actual number of segments
won't be very big. And should the need arise, allocating a bigger SGL from
slab is fast enough.

Pre-allocate only one SGL entry for protection information and switch to
runtime allocation in case that the protection information segment number
is bigger than 1. This reduces memory tied up by static command
allocations. For example, 500+ MB is saved on single lpfc HBA.

[mkp: attempted to clarify commit desc]

Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: lib/sg_pool.c: improve APIs for allocating sg pool

sg_alloc_table_chained() currently allows the caller to provide one
preallocated SGL and returns if the requested number isn't bigger than
size of that SGL. This is used to inline an SGL for an IO request.

However, scattergather code only allows that size of the 1st preallocated
SGL to be SG_CHUNK_SIZE(128). This means a substantial amount of memory
(4KB) is claimed for the SGL for each IO request. If the I/O is small, it
would be prudent to allocate a smaller SGL.

Introduce an extra parameter to sg_alloc_table_chained() and
sg_free_table_chained() for specifying size of the preallocated SGL.

Both __sg_free_table() and __sg_alloc_table() assume that each SGL has the
same size except for the last one. Change the code to allow both functions
to accept a variable size for the 1st preallocated SGL.

[mkp: attempted to clarify commit desc]

Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <bvanassche@acm.org>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Sagi Grimberg <sagi@grimberg.me>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: netdev@vger.kernel.org
Cc: linux-nvme@lists.infradead.org
Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add SPDX tags to scsi midlayer files missing licensing information

Add the default kernel GPLv2 annotation to SCSI midlayer files missing any
licensing information.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

treewide: Add SPDX license identifier for missed files

Add SPDX license identifiers to all files which:

- Have no license information of any form

- Have EXPORT_.*_SYMBOL_GPL inside which was used in the
initial scan/conversion to ignore the file

These files fall under the project license, GPL v2 only. The resulting SPDX
license identifier is:

GPL-2.0-only

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

scsi: core: set result when the command cannot be dispatched

When SCSI blk-mq is enabled, there is a bug in handling errors in
scsi_queue_rq. Specifically, the bug is not setting result field of
scsi_request correctly when the dispatch of the command has been
failed. Since the upper layer code including the sg_io ioctl expects to
receive any error status from result field of scsi_request, the error is
silently ignored and this could cause data corruptions for some
applications.

Fixes: d285203cf647 ("scsi: add support for a blk-mq based I/O path.")
Cc: <stable@vger.kernel.org>
Signed-off-by: Jaesoo Lee <jalee@purestorage.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: don't hold device refcount in IO path

scsi_device's refcount is always grabbed in IO path.

Turns out it isn't necessary, because blk_queue_cleanup() will drain any
in-flight IOs, then cancel timeout/requeue work, and SCSI's requeue_work is
canceled too in __scsi_remove_device().

Also scsi_device won't go away until blk_cleanup_queue() is done.

So don't hold the refcount in IO path, especially the refcount isn't
required in IO path since blk_queue_enter() / blk_queue_exit() is
introduced in the legacy block layer.

Cc: Dongli Zhang <dongli.zhang@oracle.com>
Cc: James Smart <james.smart@broadcom.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: linux-scsi@vger.kernel.org,
Cc: Martin K . Petersen <martin.petersen@oracle.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: James E . J . Bottomley <jejb@linux.vnet.ibm.com>,
Cc: jianchao wang <jianchao.w.wang@oracle.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Use HCTX_TYPE_DEFAULT for blk_mq_tag_set->map

Use HCTX_TYPE_DEFAULT instead of 0 to avoid hardcoding.

Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Avoid that a kernel warning appears during system resume

Since scsi_device_quiesce() skips SCSI devices that have another state than
RUNNING, OFFLINE or TRANSPORT_OFFLINE, scsi_device_resume() should not
complain about SCSI devices that have been skipped. Hence this patch. This
patch avoids that the following warning appears during resume:

WARNING: CPU: 3 PID: 1039 at blk_clear_pm_only+0x2a/0x30
CPU: 3 PID: 1039 Comm: kworker/u8:49 Not tainted 5.0.0+ #1
Hardware name: LENOVO 4180F42/4180F42, BIOS 83ET75WW (1.45 ) 05/10/2013
Workqueue: events_unbound async_run_entry_fn
RIP: 0010:blk_clear_pm_only+0x2a/0x30
Call Trace:
? scsi_device_resume+0x28/0x50
? scsi_dev_type_resume+0x2b/0x80
? async_run_entry_fn+0x2c/0xd0
? process_one_work+0x1f0/0x3f0
? worker_thread+0x28/0x3c0
? process_one_work+0x3f0/0x3f0
? kthread+0x10c/0x130
? __kthread_create_on_node+0x150/0x150
? ret_from_fork+0x1f/0x30

Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Oleksandr Natalenko <oleksandr@natalenko.name>
Cc: Martin Steigerwald <martin@lichtvoll.de>
Cc: <stable@vger.kernel.org>
Reported-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Tested-by: Jisheng Zhang <Jisheng.Zhang@synaptics.com>
Fixes: 3a0a529971ec ("block, scsi: Make SCSI quiesce and resume work reliably") # v4.15
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Also call destroy_rcu_head() for passthrough requests

cmd->rcu is initialized by scsi_initialize_rq(). For passthrough
requests, blk_get_request() calls scsi_initialize_rq(). For filesystem
requests, scsi_init_command() calls scsi_initialize_rq(). Make sure
that destroy_rcu_head() is called for passthrough requests.

Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Ewan D. Milne <emilne@redhat.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reported-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: kill command serial number

No users left, kill it.

Signed-off-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Avoid that system resume triggers a kernel warning

scsi_device_quiesce() and scsi_device_resume() are called during
system-wide suspend and resume. scsi_device_quiesce() only succeeds for
SCSI devices that are in one of the RUNNING, OFFLINE or TRANSPORT_OFFLINE
states (see also scsi_set_device_state()). This patch avoids that the
following warning is triggered when resuming a system for which quiescing a
SCSI device failed:

WARNING: CPU: 2 PID: 11303 at drivers/scsi/scsi_lib.c:2600 scsi_device_resume+0x4f/0x58
CPU: 2 PID: 11303 Comm: kworker/u8:70 Not tainted 5.0.0-rc1+ #50
Hardware name: LENOVO 80E3/Lancer 5B2, BIOS A2CN45WW(V2.13) 08/04/2016
Workqueue: events_unbound async_run_entry_fn
Call Trace:
scsi_dev_type_resume+0x2e/0x60
async_run_entry_fn+0x32/0xd8
process_one_work+0x1f4/0x420
worker_thread+0x28/0x3c0
kthread+0x118/0x130
ret_from_fork+0x22/0x40

Cc: Przemek Socha <soprwa@gmail.com>
Reported-by: Przemek Socha <soprwa@gmail.com>
Fixes: 3a0a529971ec ("block, scsi: Make SCSI quiesce and resume work reliably") # v4.15
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: reset host byte in DID_NEXUS_FAILURE case

Up to 4.12, __scsi_error_from_host_byte() would reset the host byte to
DID_OK for various cases including DID_NEXUS_FAILURE. Commit
2a842acab109 ("block: introduce new block status code type") replaced this
function with scsi_result_to_blk_status() and removed the host-byte
resetting code for the DID_NEXUS_FAILURE case. As the line
set_host_byte(cmd, DID_OK) was preserved for the other cases, I suppose
this was an editing mistake.

The fact that the host byte remains set after 4.13 is causing problems with
the sg_persist tool, which now returns success rather then exit status 24
when a RESERVATION CONFLICT error is encountered.

Fixes: 2a842acab109 "block: introduce new block status code type"
Signed-off-by: Martin Wilck <mwilck@suse.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: kill BLK_MQ_F_SG_MERGE

QUEUE_FLAG_NO_SG_MERGE has been killed, so kill BLK_MQ_F_SG_MERGE too.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Move resid from scsi_data_buffer to scsi_cmnd

This patch does not change any functionality but reduces the size of
struct scsi_cmnd.

Cc: Douglas Gilbert <dgilbert@interlog.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: stop setting up request->special

No more need in a blk-mq world where the scsi command and request are
allocated together.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: remove bidirectional command support

No real need for bidi support once the OSD code is gone.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Remove an atomic instruction from the hot path

From scsi_init_command(), a function called by scsi_mq_prep_fn():

/* zero out the cmd, except for the embedded scsi_request */
memset((char *)cmd + sizeof(cmd->req), 0,
sizeof(*cmd) - sizeof(cmd->req) + dev->host->hostt->cmd_size);

In other words, scsi_mq_prep_fn() clears scsi_cmnd.flags. Hence move the
clear_bit() call into the else branch, the only branch in which this code
is necessary.

See also commit f1342709d18a ("scsi: Do not rely on blk-mq for double
completions").

Cc: Keith Busch <keith.busch@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: communicate max segment size to the DMA mapping code

When a host driver sets a maximum segment size we should not only propagate
that setting to the block layer, which can merge segments, but also to the
DMA mapping layer which can merge segments as well.

Fixes: 50c2e9107f ("scsi: introduce a max_segment_size host_template parameters")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: remove the use_clustering flag

The same effects can be achieved by setting the dma_boundary to
PAGE_SIZE - 1 and the max_segment_size to PAGE_SIZE, so shift those
settings into the drivers. Note that in many cases the setting might
be bogus, but this keeps the status quo.

[mkp: fix myrs and myrb]

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: introduce a max_segment_size host_template parameters

This allows the host driver to indicate the maximum supported
segment size in a nice an easy way, so that the driver doesn't
have to worry about DMA-layer imposed limitations.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: flip the default on use_clustering

Most SCSI drivers want to enable "clustering", that is merging of
segments so that they might span more than a single page. Remove the
ENABLE_CLUSTERING define, and require drivers to explicitly set
DISABLE_CLUSTERING to disable this feature.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Do not rely on blk-mq for double completions

The scsi timeout error handling had been directly updating the block
layer's request state to prevent a error handling and a natural completion
from completing the same request twice. Fix this layering violation
by having scsi control the fate of its commands with scsi owned flags
rather than use blk-mq's.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

SCSI: fix queue cleanup race before queue initialization is done

c2856ae2f315d ("blk-mq: quiesce queue before freeing queue") has
already fixed this race, however the implied synchronize_rcu()
in blk_mq_quiesce_queue() can slow down LUN probe a lot, so caused
performance regression.

Then 1311326cf4755c7 ("blk-mq: avoid to synchronize rcu inside blk_cleanup_queue()")
tried to quiesce queue for avoiding unnecessary synchronize_rcu()
only when queue initialization is done, because it is usual to see
lots of inexistent LUNs which need to be probed.

However, turns out it isn't safe to quiesce queue only when queue
initialization is done. Because when one SCSI command is completed,
the user of sending command can be waken up immediately, then the
scsi device may be removed, meantime the run queue in scsi_end_request()
is still in-progress, so kernel panic can be caused.

In Red Hat QE lab, there are several reports about this kind of kernel
panic triggered during kernel booting.

This patch tries to address the issue by grabing one queue usage
counter during freeing one request and the following run queue.

Fixes: 1311326cf4755c7 ("blk-mq: avoid to synchronize rcu inside blk_cleanup_queue()")
Cc: Andrew Jones <drjones@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: linux-scsi@vger.kernel.org
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: James E.J. Bottomley <jejb@linux.vnet.ibm.com>
Cc: stable <stable@vger.kernel.org>
Cc: jianchao.wang <jianchao.w.wang@oracle.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: return blk_status_t from device handler ->prep_fn

Remove the last use of the old BLKPREP_* values, which get converted
to BLK_STS_* later anyway.

Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: return blk_status_t from scsi_init_io and ->init_command

Replace the old BLKPREP_* values with the BLK_STS_ ones that they are
converted to later anyway.

Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: clean up error handling in scsi_init_io

There is no need to call scsi_mq_free_sgtables until we have actually
allocated sgtables.

Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: push blk_status_t up into scsi_setup_{fs,scsi}_cmnd

This just moves the prep_to_mq calls up in preparation of further removal
of BLKPREP_* usage.

Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: simplify scsi_prep_state_check

Return a blk_status_t directly, and make the code a little more compact
by handling the fast path in the caller.

Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: abstract out queue map

This is in preparation for allowing multiple sets of maps per
queue, if so desired.

Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: kill off the legacy IO path

This removes the legacy (non-mq) IO path for SCSI.

Cc: linux-scsi@vger.kernel.org
Acked-by: Himanshu Madhani <himanshu.madhani@cavium.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Tested-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: provide mq_ops->busy() hook

Only the SCSI legacy path provides a way to check if target is
currently busy, provide the same for the MQ path.

Cc: linux-scsi@vger.kernel.org
Reviewed-by: Hannes Reinecke <hare@suse.com>
Tested-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Allow state transitions from OFFLINE to BLOCKED

When an RSCN gets delayed (or not being sent at all), the transport class
will detect an error, EH kicks in, and eventually will be setting the
device to offline. If we receive an RSCN after that, the device will
stay in 'offline'. This patch allows for an 'offline' to 'blocked'
transition, thereby allowing the device to become active again.

Signed-off-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block, scsi: Change the preempt-only flag into a counter

The RQF_PREEMPT flag is used for three purposes:
- In the SCSI core, for making sure that power management requests
are executed even if a device is in the "quiesced" state.
- For domain validation by SCSI drivers that use the parallel port.
- In the IDE driver, for IDE preempt requests.
Rename "preempt-only" into "pm-only" because the primary purpose of
this mode is power management. Since the power management core may
but does not have to resume a runtime suspended device before
performing system-wide suspend and since a later patch will set
"pm-only" mode as long as a block device is runtime suspended, make
it possible to set "pm-only" mode from more than one context. Since
with this change scsi_device_quiesce() is no longer idempotent, make
that function return early if it is called for a quiesced queue.

Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Cc: Jianchao Wang <jianchao.w.wang@oracle.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: remove unnecessary unlikely()

BUG_ON() already contains an unlikely(), there is no need for another one.

Signed-off-by: Igor Stoppa <igor.stoppa@huawei.com>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: "James E.J. Bottomley" <jejb@linux.vnet.ibm.com>
Cc: linux-scsi@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

Revert "scsi: core: avoid host-wide host_busy counter for scsi_mq"

This reverts commit 328728630d9f2bf14b82ca30b5e47489beefe361.

There is fundamental issue in commit 328728630d9f2bf1 (scsi: core: avoid
host-wide host_busy counter for scsi_mq) because SCSI's host busy counter
may not be same with counter of blk-mq's inflight tags, especially in case
of none io scheduler.

We may switch to other approach for addressing this scsi_mq's performance
issue, such as percpu counter or kind of ways, so revert this commit first
for fixing this kind of issue in EH path, as reported by Jens.

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Jens Axboe <axboe@kernel.dk>
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

Revert "scsi: core: fix scsi_host_queue_ready"

This reverts commit 265d59aacbce7e50bdc1f5d25033c38dd70b3767.

There is fundamental issue in commit 328728630d9f2bf1 (scsi: core: avoid
host-wide host_busy counter for scsi_mq) because SCSI's host busy counter
may not be same with counter of blk-mq's inflight tags, especially in case
of none io scheduler.

So revert this commit first.

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Jens Axboe <axboe@kernel.dk>
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: use blk_mq_run_hw_queues in scsi_kick_queue

We don't use blk-mq start/stop hw queue any more, so no reason to use
blk_mq_start_hw_queues which does clear_bit, replace it with
blk_mq_run_hw_queues.

Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Check sense buffer size at build time

To avoid introducing problems like those fixed in commit f7068114d45e
("sr: pass down correctly sized SCSI sense buffer"), this creates a macro
wrapper for scsi_execute() that verifies the size of the sense buffer
similar to what was done for command string sizes in commit 3756f6401c30
("exec: avoid gcc-8 warning for get_task_comm").

Another solution could be to add a length argument to scsi_execute(),
but this function already takes a lot of arguments and Jens was not fond
of that approach.

Additionally, this moves the SCSI_SENSE_BUFFERSIZE definition into
scsi_device.h, and removes a redundant include for scsi_device.h from
scsi_cmnd.h.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: fix scsi_host_queue_ready

328728630d9f ("scsi: avoid to hold host-wide counter of host_busy for
scsi_mq") adds one extra check on scsi_host_busy(shost) in
scsi_host_queue_ready(), which is wrong and not necessary, can causes
booting stall on LSI53c895A.

So remove the check.

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Fixes: 328728630d9f ("scsi: avoid to hold host-wide counter of host_busy for scsi_mq")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: avoid host-wide host_busy counter for scsi_mq

It isn't necessary to check the host depth in scsi_queue_rq() any more
since it has been respected by blk-mq before calling scsi_queue_rq() via
getting driver tag.

Lots of LUNs may attach to same host and per-host IOPS may reach millions,
so we should avoid expensive atomic operations on the host-wide counter in
the IO path.

This patch implements scsi_host_busy() via blk_mq_tagset_busy_iter() for
reading the count of busy IOs for scsi_mq.

It is observed that IOPS is increased by 15% in IO test on scsi_debug (32
LUNs, 32 submit queues, 1024 can_queue, libaio/dio) in a dual-socket
system.

[mkp: clarified commit message]

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: check for equality of result byte values

When evaluating a SCSI command's result using the field access macros,
check for equality of the fields and not if a specific bit is set.

This is a preparation patch, for reworking the results field in the
SCSI command.

Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: scsi_io_completion convert BUGs to WARNs

The scsi_io_completion function contains three BUG() and BUG_ON() calls.
Replace them with WARN variants.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: scsi_io_completion hints on fastpath

Add likely() and unlikely() hints to conditionals on or near the fastpath.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add scsi_io_completion_reprep helper

Since the action "reprep" is called from two places, rather than repeat the
code, make a new scsi_io_completion helper with "reprep" as its suffix.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add scsi_io_completion_action helper

Place scsi_io_completion()'s complex error processing associated with a
local enumeration into a static helper function. That enumeration's values
start with "ACTION_" so use the suffix "_action" in the helper function's
name.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: add scsi_io_completion_nz_result function

Break out several intertwined paths when cmd->result is non zero and place
them in the scsi_io_completion_nz_result helper function. The logic is not
changed.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: scsi_io_completion: rename variables

Change and add some variable names, adjust some associated comments for
clarity. Correct some misleading comments.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: scsi_io_completion: comment on end_request return

scsi_end_request() is called multiple times from scsi_io_completion() which
branches on its bool returned value. Add comment before the static
definition of scsi_end_request() about the meaning of that return.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: sanitize++ in progress

Commit 505aa4b6a883 ("scsi: sd: Defer spinning up drive while SANITIZE is
in progress") may not be sufficient, especially if the SCSI SANITIZE
command is sent via the bsg or sg pass-throughs, since they don't use the
sd driver.

Add "Sanitize in progress" plus some other recent "... in progress"
additional sense codes into the scsi mid-level so they are treated in a
similar fashion to "Format in progress".

[mkp: checkpatch]

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: consistently use GFP_NOIO instead of __GFP_NORECLAIM

Same numerical value (for now at least), but a much better documentation
of intent.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: sanitize blk_get_request calling conventions

Switch everyone to blk_get_request_flags, and then rename
blk_get_request_flags to blk_get_request.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: reduce use of block bounce buffers

We can rely on the dma-mapping code to handle any DMA limits that is
bigger than the ISA DMA mask for us (either using an iommu or swiotlb),
so remove setting the block layer bounce limit for anything but the
unchecked_isa_dma case, or the bouncing for highmem pages.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Make scsi_result_to_blk_status() recognize CONDITION MET

Ensure that CONDITION MET and other non-zero status values that indicate
success are translated into BLK_STS_OK.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Douglas Gilbert <dgilbert@interlog.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Lee Duncan <lduncan@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Rename __scsi_error_from_host_byte() into scsi_result_to_blk_status()

Since the next patch will modify this function such that it checks more than
just the host byte of the SCSI result, rename __scsi_error_from_host_byte()
into scsi_result_to_blk_status(). This patch does not change any
functionality.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Douglas Gilbert <dgilbert@interlog.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Lee Duncan <lduncan@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

Revert "scsi: core: return BLK_STS_OK for DID_OK in __scsi_error_from_host_byte()"

The description of commit e39a97353e53 is wrong: it mentions that commit
2a842acab109 introduced a bug in __scsi_error_from_host_byte() although that
commit did not change the behavior of that function. Additionally, commit
e39a97353e53 introduced a bug: it causes commands that fail with
hostbyte=DID_OK and driverbyte=DRIVER_SENSE to be completed with
BLK_STS_OK. Hence revert that commit.

Fixes: e39a97353e53 ("scsi: core: return BLK_STS_OK for DID_OK in __scsi_error_from_host_byte()")
Reported-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Douglas Gilbert <dgilbert@interlog.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Lee Duncan <lduncan@suse.com>
Cc: stable@vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

bsg: split handling of SCSI CDBs vs transport requeues

The current BSG design tries to shoe-horn the transport-specific
passthrough commands into the overall framework for SCSI passthrough
requests. This has a couple problems:

- each passthrough queue has to set the QUEUE_FLAG_SCSI_PASSTHROUGH flag
despite not dealing with SCSI commands at all. Because of that these
queues could also incorrectly accept SCSI commands from in-kernel
users or through the legacy SCSI_IOCTL_SEND_COMMAND ioctl.
- the real SCSI bsg queues also incorrectly accept bsg requests of the
BSG_SUB_PROTOCOL_SCSI_TRANSPORT type
- the bsg transport code is almost unredable because it tries to reuse
different SCSI concepts for its own purpose.

This patch instead adds a new bsg_ops structure to handle the two cases
differently, and thus solves all of the above problems. Another side
effect is that the bsg-lib queues also don't need to embedd a
struct scsi_request anymore.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Make SCSI Status CONDITION MET equivalent to GOOD

The SCSI PRE-FETCH (10 or 16) command is present both on hard disks
and some SSDs. It is useful when the address of the next block(s) to
be read is known but it is not following the LBA of the current READ
(so read-ahead won't help). It returns two "good" SCSI Status values.
If the requested blocks have fitted (or will most likely fit (when
the IMMED bit is set)) into the disk's cache, it returns CONDITION
MET. If it didn't (or will not) fit then it returns GOOD status.

The goal of this patch is to stop the SCSI subsystem treating the
CONDITION MET SCSI status as an error. The current state makes the
PRE-FETCH command effectively unusable via pass-throughs.

Signed-off-by: Douglas Gilbert <dgilbert@interlog.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Use blk_queue_flag_*() in drivers instead of queue_flag_*()

This patch has been generated as follows:

for verb in set_unlocked clear_unlocked set clear; do
replace-in-files queue_flag_${verb} blk_queue_flag_${verb%_unlocked} \
$(git grep -lw queue_flag_${verb} drivers block/bsg*)
done

Except for protecting all queue flag changes with the queue lock
this patch does not change any functionality.

Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Shaohua Li <shli@fb.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: use blk_mq_requeue_request in __scsi_queue_insert

In scsi core, __scsi_queue_insert should just put request back on the
queue and retry using the same command as before. However, for blk-mq,
scsi_mq_requeue_cmd is employed here which will unprepare the
request. To align with the semantics of __scsi_queue_insert, use
blk_mq_requeue_request with kick_requeue_list == true and put the
reference of scsi_device.

Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: return BLK_STS_OK for DID_OK in __scsi_error_from_host_byte()

When converting __scsi_error_from_host_byte() to BLK_STS error codes the
case DID_OK was forgotten, resulting in it always returning an error.

Fixes: 2a842acab109 ("block: introduce new block status code type")
Cc: Doug Gilbert <dgilbert@interlog.com>
Signed-off-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Douglas Gilbert <dgilbert@interlog.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Avoid that ATA error handling can trigger a kernel hang or oops

Avoid that the recently introduced call_rcu() call in the SCSI core
triggers a double call_rcu() call.

Reported-by: Natanael Copa <ncopa@alpinelinux.org>
Reported-by: Damien Le Moal <damien.lemoal@wdc.com>
References: https://bugzilla.kernel.org/show_bug.cgi?id=198861
Fixes: 3bd6f43f5cb3 ("scsi: core: Ensure that the SCSI error handler gets woken up")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Damien Le Moal <damien.lemoal@wdc.com>
Tested-by: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Natanael Copa <ncopa@alpinelinux.org>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Alexandre Oliva <oliva@gnu.org>
Cc: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Add 'lock' as third argument to blk_alloc_queue_node()

This patch does not change any functionality.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Philipp Reisner <philipp.reisner@linbit.com>
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Cc: Kees Cook <keescook@chromium.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: core: Reduce number of scsi_test_unit_ready() retries

Make scsi_test_unit_ready() send at most as many TURs as specified in
the 'retries' argument instead of retries * (retries + 1) / 2.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

blk-mq: introduce BLK_STS_DEV_RESOURCE

This status is returned from driver to block layer if device related
resource is unavailable, but driver can guarantee that IO dispatch
will be triggered in future when the resource is available.

Convert some drivers to return BLK_STS_DEV_RESOURCE. Also, if driver
returns BLK_STS_RESOURCE and SCHED_RESTART is set, rerun queue after
a delay (BLK_MQ_DELAY_QUEUE) to avoid IO stalls. BLK_MQ_DELAY_QUEUE is
3 ms because both scsi-mq and nvmefc are using that magic value.

If a driver can make sure there is in-flight IO, it is safe to return
BLK_STS_DEV_RESOURCE because:

1) If all in-flight IOs complete before examining SCHED_RESTART in
blk_mq_dispatch_rq_list(), SCHED_RESTART must be cleared, so queue
is run immediately in this case by blk_mq_dispatch_rq_list();

2) if there is any in-flight IO after/when examining SCHED_RESTART
in blk_mq_dispatch_rq_list():
- if SCHED_RESTART isn't set, queue is run immediately as handled in 1)
- otherwise, this request will be dispatched after any in-flight IO is
completed via blk_mq_sched_restart()

3) if SCHED_RESTART is set concurently in context because of
BLK_STS_RESOURCE, blk_mq_delay_run_hw_queue() will cover the above two
cases and make sure IO hang can be avoided.

One invariant is that queue will be rerun if SCHED_RESTART is set.

Suggested-by: Jens Axboe <axboe@kernel.dk>
Tested-by: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: Define usercopy region in scsi_sense_cache slab cache

SCSI sense buffers, stored in struct scsi_cmnd.sense and therefore
contained in the scsi_sense_cache slab cache, need to be copied to/from
userspace.

cache object allocation:
drivers/scsi/scsi_lib.c:
scsi_select_sense_cache(...):
return ... ? scsi_sense_isadma_cache : scsi_sense_cache

scsi_alloc_sense_buffer(...):
return kmem_cache_alloc_node(scsi_select_sense_cache(), ...);

scsi_init_request(...):
...
cmd->sense_buffer = scsi_alloc_sense_buffer(...);
...
cmd->req.sense = cmd->sense_buffer

example usage trace:

block/scsi_ioctl.c:
(inline from sg_io)
blk_complete_sghdr_rq(...):
struct scsi_request *req = scsi_req(rq);
...
copy_to_user(..., req->sense, len)

scsi_cmd_ioctl(...):
sg_io(...);

In support of usercopy hardening, this patch defines a region in
the scsi_sense_cache slab cache in which userspace copy operations
are allowed.

This region is known as the slab cache's usercopy region. Slab caches
can now check that each dynamically sized copy operation involving
cache-managed memory falls entirely within the slab's usercopy region.

Signed-off-by: David Windsor <dave@nullcore.net>
[kees: adjust commit log, provide usage trace]
Cc: "James E.J. Bottomley" <jejb@linux.vnet.ibm.com>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: linux-scsi@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>

scsi: core: Change third __scsi_queue_insert() argument from int to bool

This patch does not change any functionality but makes the SCSI core
source code slightly easier to read.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Unexport scsi_initialize_rq()

Commit 651a01364994 ("scsi: scsi_transport_sas: switch to bsg-lib for
SMP passthrough") removed the only call to scsi_initialize_rq() from
outside the SCSI core. Hence unexport scsi_initialize_rq().

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: Ensure that the SCSI error handler gets woken up

If scsi_eh_scmd_add() is called concurrently with
scsi_host_queue_ready() while shost->host_blocked > 0 then it can
happen that neither function wakes up the SCSI error handler. Fix
this by making every function that decreases the host_busy counter
wake up the error handler if necessary and by protecting the
host_failed checks with the SCSI host lock.

Reported-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
References: https://marc.info/?l=linux-kernel&m=150461610630736
Fixes: commit 746650160866 ("scsi: convert host_busy to atomic_t")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Tested-by: Stuart Hayes <stuart.w.hayes@gmail.com>
Cc: Konstantin Khorenko <khorenko@virtuozzo.com>
Cc: Stuart Hayes <stuart.w.hayes@gmail.com>
Cc: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: core: run queue if SCSI device queue isn't ready and queue is idle

Before commit 0df21c86bdbf ("scsi: implement .get_budget and .put_budget
for blk-mq"), we run queue after 3ms if queue is idle and SCSI device
queue isn't ready, which is done in handling BLK_STS_RESOURCE. After
commit 0df21c86bdbf is introduced, queue won't be run any more under
this situation.

IO hang is observed when timeout happened, and this patch fixes the IO
hang issue by running queue after delay in scsi_dev_queue_ready, just
like non-mq. This issue can be triggered by the following script[1].

There is another issue which can be covered by running idle queue: when
.get_budget() is called on request coming from hctx->dispatch_list, if
one request just completes during .get_budget(), we can't depend on
SCSI's restart to make progress any more. This patch fixes the race too.

With this patch, we basically recover to previous behaviour (before
commit 0df21c86bdbf) of handling idle queue when running out of
resource.

[1] script for test/verify SCSI timeout
rmmod scsi_debug
modprobe scsi_debug max_queue=1

DEVICE=`ls -d /sys/bus/pseudo/drivers/scsi_debug/adapter*/host*/target*/*/block/* | head -1 | xargs basename`
DISK_DIR=`ls -d /sys/block/$DEVICE/device/scsi_disk/*`

echo "using scsi device $DEVICE"
echo "-1" >/sys/bus/pseudo/drivers/scsi_debug/every_nth
echo "temporary write through" >$DISK_DIR/cache_type
echo "128" >/sys/bus/pseudo/drivers/scsi_debug/opts
echo none > /sys/block/$DEVICE/queue/scheduler
dd if=/dev/$DEVICE of=/dev/null bs=1M iflag=direct count=1 &
sleep 5
echo "0" >/sys/bus/pseudo/drivers/scsi_debug/opts
wait
echo "SUCCESS"

Fixes: 0df21c86bdbf ("scsi: implement .get_budget and .put_budget for blk-mq")
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com>
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: use dma_get_cache_alignment() as minimum DMA alignment

In non-coherent DMA mode, kernel uses cache flushing operations to
maintain I/O coherency, so scsi's block queue should be aligned to the
value returned by dma_get_cache_alignment(). Otherwise, If a DMA buffer
and a kernel structure share a same cache line, and if the kernel
structure has dirty data, cache_invalidate (no writeback) will cause
data corruption.

Cc: stable@vger.kernel.org
Signed-off-by: Huacai Chen <chenhc@lemote.com>
[hch: rebased and updated the comment and changelog]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block, scsi: Make SCSI quiesce and resume work reliably

The contexts from which a SCSI device can be quiesced or resumed are:
* Writing into /sys/class/scsi_device/*/device/state.
* SCSI parallel (SPI) domain validation.
* The SCSI device power management methods. See also scsi_bus_pm_ops.

It is essential during suspend and resume that neither the filesystem
state nor the filesystem metadata in RAM changes. This is why while
the hibernation image is being written or restored that SCSI devices
are quiesced. The SCSI core quiesces devices through scsi_device_quiesce()
and scsi_device_resume(). In the SDEV_QUIESCE state execution of
non-preempt requests is deferred. This is realized by returning
BLKPREP_DEFER from inside scsi_prep_state_check() for quiesced SCSI
devices. Avoid that a full queue prevents power management requests
to be submitted by deferring allocation of non-preempt requests for
devices in the quiesced state. This patch has been tested by running
the following commands and by verifying that after each resume the
fio job was still running:

for ((i=0; i<10; i++)); do
(
cd /sys/block/md0/md &&
while true; do
[ "$(<sync_action)" = "idle" ] && echo check > sync_action
sleep 1
done
) &
pids=($!)
for d in /sys/class/block/sd*[a-z]; do
bdev=${d#/sys/class/block/}
hcil=$(readlink "$d/device")
hcil=${hcil#../../../}
echo 4 > "$d/queue/nr_requests"
echo 1 > "/sys/class/scsi_device/$hcil/device/queue_depth"
fio --name="$bdev" --filename="/dev/$bdev" --buffered=0 --bs=512 \
--rw=randread --ioengine=libaio --numjobs=4 --iodepth=16 \
--iodepth_batch=1 --thread --loops=$((2**31)) &
pids+=($!)
done
sleep 1
echo "$(date) Hibernating ..." >>hibernate-test-log.txt
systemctl hibernate
sleep 10
kill "${pids[@]}"
echo idle > /sys/block/md0/md/sync_action
wait
echo "$(date) Done." >>hibernate-test-log.txt
done

Reported-by: Oleksandr Natalenko <oleksandr@natalenko.name>
References: "I/O hangs after resuming from suspend-to-ram" (https://marc.info/?l=linux-block&m=150340235201348).
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Tested-by: Martin Steigerwald <martin@lichtvoll.de>
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

ide, scsi: Tell the block layer at request allocation time about preempt requests

Convert blk_get_request(q, op, __GFP_RECLAIM) into
blk_get_request_flags(q, op, BLK_MQ_PREEMPT). This patch does not
change any functionality.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Tested-by: Martin Steigerwald <martin@lichtvoll.de>
Acked-by: David S. Miller <davem@davemloft.net> [ for IDE ]
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

Revert "scsi: make 'state' device attribute pollable"

This reverts commit 8a97712e5314aefe16b3ffb4583a34deaa49de04.

This commit added a call to sysfs_notify() from within
scsi_device_set_state(), which in turn turns out to make libata very
unhappy, because ata_eh_detach_dev() does

spin_lock_irqsave(ap->lock, flags);
..
if (ata_scsi_offline_dev(dev)) {
dev->flags |= ATA_DFLAG_DETACHED;
ap->pflags |= ATA_PFLAG_SCSI_HOTPLUG;
}

and ata_scsi_offline_dev() then does that scsi_device_set_state() to set
it offline.

So now we called sysfs_notify() from within a spinlocked region, which
really doesn't work. The 0day robot reported this as:

BUG: sleeping function called from invalid context at kernel/locking/mutex.c:238

because sysfs_notify() ends up calling kernfs_find_and_get_ns() which
then does mutex_lock(&kernfs_mutex)..

The pollability of the device state isn't critical, so revert this all
for now, and maybe we'll do it differently in the future.

Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
Acked-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

blk-mq: don't handle failure in .get_budget

It is enough to just check if we can get the budget via .get_budget().
And we don't need to deal with device state change in .get_budget().

For SCSI, one issue to be fixed is that we have to call
scsi_mq_uninit_cmd() to free allocated ressources if SCSI device fails
to handle the request. And it isn't enough to simply call
blk_mq_end_request() to do that if this request is marked as
RQF_DONTPREP.

Fixes: 0df21c86bdbf(scsi: implement .get_budget and .put_budget for blk-mq)
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

SCSI: don't get target/host busy_count in scsi_mq_get_budget()

It is very expensive to atomic_inc/atomic_dec the host wide counter of
host->busy_count, and it should have been avoided via blk-mq's mechanism
of getting driver tag, which uses the more efficient way of sbitmap queue.

Also we don't check atomic_read(&sdev->device_busy) in scsi_mq_get_budget()
and don't run queue if the counter becomes zero, so IO hang may be caused
if all requests are completed just before the current SCSI device
is added to shost->starved_list.

Fixes: 0df21c86bdbf(scsi: implement .get_budget and .put_budget for blk-mq)
Reported-by: Bart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: implement .get_budget and .put_budget for blk-mq

We need to tell blk-mq to reserve resources before queuing one request,
so implement these two callbacks. Then blk-mq can avoid to dequeue
request too early, and IO merging can be improved a lot.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: allow passing in null rq to scsi_prep_state_check()

In the following patch, we will implement scsi_get_budget()
which need to call scsi_prep_state_check() when rq isn't
dequeued yet.

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: Suppress a kernel warning in case the prep function returns BLKPREP_DEFER

The legacy block layer handles requests as follows:
- If the prep function returns BLKPREP_OK, let blk_peek_request()
return the pointer to that request.
- If the prep function returns BLKPREP_DEFER, keep the RQF_STARTED
flag and retry calling the prep function later.
- If the prep function returns BLKPREP_KILL or BLKPREP_INVALID, end
the request.

In none of these cases it is correct to clear the SCMD_INITIALIZED
flag from inside scsi_prep_fn(). Since scsi_prep_fn() already
guarantees that scsi_init_command() will be called once even if
scsi_prep_fn() is called multiple times, remove the code that clears
SCMD_INITIALIZED from scsi_prep_fn().

The scsi-mq code handles requests as follows:
- If scsi_mq_prep_fn() returns BLKPREP_OK, set the RQF_DONTPREP flag
and submit the request to the SCSI LLD.
- If scsi_mq_prep_fn() returns BLKPREP_DEFER, call
blk_mq_delay_run_hw_queue() and return BLK_STS_RESOURCE.
- If the prep function returns BLKPREP_KILL or BLKPREP_INVALID, call
scsi_mq_uninit_cmd() and let the blk-mq core end the request.

In none of these cases scsi_mq_prep_fn() should clear the
SCMD_INITIALIZED flag. Hence remove the code from scsi_mq_prep_fn()
function that clears that flag.

This patch avoids that the following warning is triggered when using
the legacy block layer:

------------[ cut here ]------------
WARNING: CPU: 1 PID: 4198 at drivers/scsi/scsi_lib.c:654 scsi_end_request+0x1de/0x220
CPU: 1 PID: 4198 Comm: mkfs.f2fs Not tainted 4.14.0-rc5+ #1
task: ffff91c147a4b800 task.stack: ffffb282c37b8000
RIP: 0010:scsi_end_request+0x1de/0x220
Call Trace:
<IRQ>
scsi_io_completion+0x204/0x5e0
scsi_finish_command+0xce/0xe0
scsi_softirq_done+0x126/0x130
blk_done_softirq+0x6e/0x80
__do_softirq+0xcf/0x2a8
irq_exit+0xab/0xb0
do_IRQ+0x7b/0xc0
common_interrupt+0x90/0x90
</IRQ>
RIP: 0010:_raw_spin_unlock_irqrestore+0x9/0x10
__test_set_page_writeback+0xc7/0x2c0
__block_write_full_page+0x158/0x3b0
block_write_full_page+0xc4/0xd0
blkdev_writepage+0x13/0x20
__writepage+0x12/0x40
write_cache_pages+0x204/0x500
generic_writepages+0x48/0x70
blkdev_writepages+0x9/0x10
do_writepages+0x34/0xc0
__filemap_fdatawrite_range+0x6c/0x90
file_write_and_wait_range+0x31/0x90
blkdev_fsync+0x16/0x40
vfs_fsync_range+0x44/0xa0
do_fsync+0x38/0x60
SyS_fsync+0xb/0x10
entry_SYSCALL_64_fastpath+0x13/0x94
---[ end trace 86e8ef85a4a6c1d1 ]---

Fixes: commit 64104f703212 ("scsi: Call scsi_initialize_rq() for filesystem requests")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: scsi_error: Handle power-on reset unit attention

As per SAM there is a status precedence, with any sense code 29/XX
taking second place just after an ACA ACTIVE status. Additionally, each
target might prefer to not queue any unit attention conditions, but just
report one. Due to the above, this will be that one with the highest
precedence. This results in the sense code 29/XX effectively
overwriting any other unit attention. Hence we should report the
power-on reset to userland so that it can take appropriate action.

Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: sd_zbc: Fix comments and indentation

Fix comments style (use kernel-doc style) and content to clarify some
functions. Also fix some functions signature indentation and remove a
useless blank line in sd_zbc_read_zones().

No functional change is introduced by this patch.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: scsi-mq: Always unprepare before requeuing a request

One of the two scsi-mq functions that requeue a request unprepares a
request before requeueing (scsi_io_completion()) but the other function
not (__scsi_queue_insert()). Make sure that a request is unprepared
before requeuing it.

Fixes: commit d285203cf647 ("scsi: add support for a blk-mq based I/O path.")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Damien Le Moal <damien.lemoal@wdc.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Improve requeuing behavior

Requests are unprepared and reprepared when being requeued. Avoid that
requeuing resets .jiffies_at_alloc and .retries by initializing these
two member variables from inside scsi_initialize_rq() and by preserving
both member variables when preparing a request. This patch affects the
requeuing behavior of both the legacy scsi and the scsi-mq code paths.

Reported-by: Brian King <brking@linux.vnet.ibm.com>
References: https://lkml.org/lkml/2017/8/18/923 ("Re: [BUG][bisected 270065e] linux-next fails to boot on powerpc")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Brian King <brking@linux.vnet.ibm.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Call scsi_initialize_rq() for filesystem requests

If a pass-through request is submitted then blk_get_request()
initializes that request by calling scsi_initialize_rq(). Also call this
function for filesystem requests. Introduce CMD_INITIALIZED to keep
track of whether or not a request has already been initialized.

Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Brian King <brking@linux.vnet.ibm.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Rework handling of scsi_device.vpd_pg8[03]

Introduce struct scsi_vpd for the VPD page length, data and the RCU head
that will be used to free the VPD data. Use kfree_rcu() instead of
kfree() to free VPD data. Move the VPD buffer pointer check inside the
RCU read lock in the sysfs code. Only annotate pointers that are shared
across threads with __rcu. Use rcu_dereference() when dereferencing an
RCU pointer. This patch suppresses about twenty sparse complaints about
the vpd_pg8[03] pointers. This patch also fixes a race condition, namely
that updating of the VPD pointers and length variables in struct
scsi_device was not atomic with reference to the code reading these
variables. See also "Does the update code tolerate concurrent accesses?"
in Documentation/RCU/checklist.txt.

Fixes: commit 09e2b0b14690 ("scsi: rescan VPD attributes")
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Acked-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Shane Seymour <shane.seymour@hpe.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Shane Seymour <shane.seymour@hpe.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Fix the kerneldoc for scsi_initialize_rq()

The kerneldoc comment for scsi_initialize_rq() neglected to document the
"rq" parameter, leading to this docs build warning:

./drivers/scsi/scsi_lib.c:1116: warning: No description found for parameter 'rq'

Document the parameter and make the build slightly quieter.

[mkp: used wording suggested by Bart]

Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: fix comment in scsi_device_set_state()

The function returns '0' if successful; with the original comment
the function doesn't have a way to indicate success ...

Signed-off-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Bart van Assche <bvanassche@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Use blk_mq_rq_to_pdu() to convert a request to a SCSI command pointer

Since commit e9c787e65c0c ("scsi: allocate scsi_cmnd structures as
part of struct request") struct request and struct scsi_cmnd are
adjacent. This means that there is now an alternative to reading
req->special to convert a pointer to a prepared request into a
SCSI command pointer, namely by using blk_mq_rq_to_pdu(). Make
this change where appropriate. Although this patch does not
change any functionality, it slightly improves performance and
slightly improves readability.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Document which queue type a function is intended for

Rename several functions to make it easy to see which queue type a
function is intended for.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: make 'state' device attribute pollable

While the 'state' attribute can (and will) change occasionally,
calling 'poll()' or 'select()' on it fails as sysfs is never
notified that the state has changed.
With this patch calling 'poll()' or 'select()' will work
properly.

Signed-off-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: scsi_lib: rework scsi_internal_device_unblock_nowait()

Rework scsi_internal_device_unblock_nowait() into using a switch
statement. No functional changes.

Signed-off-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

block: Change argument type of scsi_req_init()

Since scsi_req_init() works on a struct scsi_request, change the
argument type into struct scsi_request *.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

block: Make most scsi_req_init() calls implicit

Instead of explicitly calling scsi_req_init() after blk_get_request(),
call that function from inside blk_get_request(). Add an
.initialize_rq_fn() callback function to the block drivers that need
it. Merge the IDE .init_rq_fn() function into .initialize_rq_fn()
because it is too small to keep it as a separate function. Keep the
scsi_req_init() call in ide_prep_sense() because it follows a
blk_rq_init() call.

References: commit 82ed4db499b8 ("block: split scsi_request out of struct request")
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Omar Sandoval <osandov@fb.com>
Cc: Nicholas Bellinger <nab@linux-iscsi.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

blk-mq: use the introduced blk_mq_unquiesce_queue()

blk_mq_unquiesce_queue() is used for unquiescing the
queue explicitly, so replace blk_mq_start_stopped_hw_queues()
with it.

For the scsi part, this patch takes Bart's suggestion to
switch to block quiesce/unquiesce API completely.

Cc: linux-nvme@lists.infradead.org
Cc: linux-scsi@vger.kernel.org
Cc: dm-devel@redhat.com
Reviewed-by: Bart Van Assche <Bart.VanAssche@sandisk.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

scsi: Make scsi_mq_prep_fn() call scsi_init_command()

This patch reduces code duplication. There are two functional changes in
this patch:

- It causes scsi_mq_prep_fn() to clear driver-private command data, just
like the already upstream commit 1bad6c4a57ef ("scsi: zero per-cmd
private driver data for each MQ I/O").

- The initialization of .prot_sdb is moved from scsi_mq_prep_fn() into
scsi_init_request().

[mkp: applied by hand]

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Introduce scsi_mq_sgl_size()

This patch does not change any functionality but makes the next patch
easier to read.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Only add commands to the device command list if required by the LLD

Just like for the scsi-mq code path, in the single queue SCSI code path
only add commands to the per-device command list if required by the SCSI
LLD. This patch will make it easier to merge the single-queue and
multiqueue command initialization code.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

scsi: Make __scsi_remove_device go straight from BLOCKED to DEL

If a device is blocked, make __scsi_remove_device() cause it to
transition to the DEL state. This means that all the commands issued in
.shutdown() will error in the mid-layer, thus making the removal proceed
without being stopped.

This patch is a slightly modified version of a patch from James
Bottomley. This patch avoids that the following lockup occurs:

Call Trace:
schedule+0x35/0x80
schedule_timeout+0x237/0x2d0
io_schedule_timeout+0xa6/0x110
wait_for_completion_io+0xa3/0x110
blk_execute_rq+0xdf/0x120
scsi_execute+0xce/0x150 [scsi_mod]
scsi_execute_req_flags+0x8f/0xf0 [scsi_mod]
sd_sync_cache+0xa9/0x190 [sd_mod]
sd_shutdown+0x6a/0x100 [sd_mod]
sd_remove+0x64/0xc0 [sd_mod]
__device_release_driver+0x8d/0x120
device_release_driver+0x1e/0x30
bus_remove_device+0xf9/0x170
device_del+0x127/0x240
__scsi_remove_device+0xc1/0xd0 [scsi_mod]
scsi_forget_host+0x57/0x60 [scsi_mod]
scsi_remove_host+0x72/0x110 [scsi_mod]
srp_remove_work+0x8b/0x200 [ib_srp]

Reported-by: Israel Rukshin <israelr@mellanox.com>
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Israel Rukshin <israelr@mellanox.com>
Cc: Max Gurtovoy <maxg@mellanox.com>
Cc: Benjamin Block <bblock@linux.vnet.ibm.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>