delta/linux.git/include/linux/device-mapper.h, branch proc-cmdline

Merge tag 'for-4.17/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm

2018-04-06T18:50:19+00:00

Pull device mapper updates from Mike Snitzer:

- DM core passthrough ioctl fix to retain reference to DM table, and
that table's block devices, while issuing the ioctl to one of those
block devices.

- DM core passthrough ioctl fix to _not_ override the fmode_t used to
issue the ioctl. Overriding by using the fmode_t that the block
device was originally open with during DM table load is a liability.

- Add DM core support for secure erase forwarding and update the DM
linear and DM striped targets to support them.

- A DM core 4.16 stable fix to allow abnormal IO (e.g. discard, write
same, write zeroes) for targets that make use of the non-splitting IO
variant (as is done for multipath or thinp when layered directly on
NVMe).

- Allow DM targets to return a payload in response to a DM message that
they are sent. This is useful for DM targets that would like to
provide statistics data in response to DM messages.

- Update DM bufio to support non-power-of-2 block sizes. Numerous other
related changes prepare the DM bufio code for this support.

- Fix DM crypt to use a bounded amount of memory across the entire
system. This is to avoid OOM that can otherwise occur in response to
certain pathological IO workloads (e.g. discarding a large DM crypt
device).

- Add a 'check_at_most_once' feature to the DM verity target to allow
verity to be used on mobile devices that have very limited resources.

- Fix the DM integrity target to fail early if a keyed algorithm (e.g.
HMAC) is to be used but the key isn't set.

- Add non-power-of-2 support to the DM unstripe target.

- Eliminate the use of a Variable Length Array in the DM stripe target.

- Update the DM log-writes target to record metadata (REQ_META flag).

- DM raid fixes for its nosync status and some variable range issues.

* tag 'for-4.17/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm: (28 commits)
dm: remove fmode_t argument from .prepare_ioctl hook
dm: hold DM table for duration of ioctl rather than use blkdev_get
dm raid: fix parse_raid_params() variable range issue
dm verity: make verity_for_io_block static
dm verity: add 'check_at_most_once' option to only validate hashes once
dm bufio: don't embed a bio in the dm_buffer structure
dm bufio: support non-power-of-two block sizes
dm bufio: use slab cache for dm_buffer structure allocations
dm bufio: reorder fields in dm_buffer structure
dm bufio: relax alignment constraint on slab cache
dm bufio: remove code that merges slab caches
dm bufio: get rid of slab cache name allocations
dm bufio: move dm-bufio.h to include/linux/
dm bufio: delete outdated comment
dm: add support for secure erase forwarding
dm: backfill abnormal IO support to non-splitting IO submission
dm raid: fix nosync status
dm mpath: use DM_MAPIO_SUBMITTED instead of magic number 0 in process_queued_bios()
dm stripe: get rid of a Variable Length Array (VLA)
dm log writes: record metadata flag for better flags record
...

dm: remove fmode_t argument from .prepare_ioctl hook

2018-04-04T16:12:39+00:00

Use the fmode_t that is passed to dm_blk_ioctl() rather than
inconsistently (varies across targets) drop it on the floor by
overriding it with the fmode_t stored in 'struct dm_dev'.

All the persistent reservation functions weren't using the fmode_t they
got back from .prepare_ioctl so remove them.

Signed-off-by: Mike Snitzer

dm: add support for secure erase forwarding

2018-04-03T19:04:21+00:00

Set QUEUE_FLAG_SECERASE in DM device's queue_flags if a DM table's
data devices support secure erase.

Also, add support for secure erase to both the linear and striped
targets.

Signed-off-by: Denis Semakin 
Signed-off-by: Mike Snitzer

dm: allow targets to return output from messages they are sent

2018-04-03T19:04:10+00:00

Could be useful for a target to return stats or other information.
If a target does DMEMIT() anything to @result from its .message method
then it must return 1 to the caller.

Signed-off-By: Mike Snitzer

block: Move SECTOR_SIZE and SECTOR_SHIFT definitions into

2018-03-17T20:45:23+00:00

It happens often while I'm preparing a patch for a block driver that
I'm wondering: is a definition of SECTOR_SIZE and/or SECTOR_SHIFT
available for this driver? Do I have to introduce definitions of these
constants before I can use these constants? To avoid this confusion,
move the existing definitions of SECTOR_SIZE and SECTOR_SHIFT into the
 header file such that these become available for all
block drivers. Make the SECTOR_SIZE definition in the uapi msdos_fs.h
header file conditional to avoid that including that header file after
 causes the compiler to complain about a SECTOR_SIZE
redefinition.

Note: the SECTOR_SIZE / SECTOR_SHIFT / SECTOR_BITS definitions have
not been removed from uapi header files nor from NAND drivers in
which these constants are used for another purpose than converting
block layer offsets and sizes into a number of sectors.

Cc: David S. Miller 
Cc: Mike Snitzer 
Cc: Dan Williams 
Cc: Minchan Kim 
Cc: Nitin Gupta 
Reviewed-by: Sergey Senozhatsky 
Reviewed-by: Christoph Hellwig 
Reviewed-by: Johannes Thumshirn 
Reviewed-by: Martin K. Petersen 
Signed-off-by: Bart Van Assche 
Signed-off-by: Jens Axboe

dm mpath: delay the retry of a request if the target responded as busy

2018-01-29T18:44:54+00:00

Add DM_ENDIO_DELAY_REQUEUE to allow request-based multipath's
multipath_end_io() to instruct dm-rq.c:dm_done() to delay a requeue.
This is beneficial to do if BLK_STS_RESOURCE is returned from the target
(because target is busy).

Relative to blk-mq: kick the hw queues via blk_mq_requeue_work(),
indirectly from dm-rq.c:__dm_mq_kick_requeue_list(), after a delay.

For old .request_fn: use blk_delay_queue().

bio-based multipath doesn't have feature parity with request-based for
retryable error requeues; that is something that'll need fixing in the
future.

Suggested-by: Bart Van Assche 
Signed-off-by: Mike Snitzer 
Acked-by: Bart Van Assche 
[as interpreted from Bart's "... patch looks fine to me."]

dm: move dm_table_destroy() to same header as dm_table_create()

2018-01-17T14:16:06+00:00

If anyone is going to use dm_table_create(), they probably should be
able to use dm_table_destroy() too. Move the dm_table_destroy()
definition outside the private header, near dm_table_create()

Signed-off-by: Brian Norris 
Signed-off-by: Mike Snitzer

dm: introduce DM_TYPE_NVME_BIO_BASED

2017-12-20T15:51:10+00:00

If dm_table_determine_type() establishes DM_TYPE_NVME_BIO_BASED then
all devices in the DM table do not support partial completions.  Also,
the table has a single immutable target that doesn't require DM core to
split bios.

This will enable adding NVMe optimizations to bio-based DM.

Signed-off-by: Mike Snitzer

dm: improve performance by moving dm_io structure to per-bio-data

2017-12-17T01:43:13+00:00

Eliminates need for a separate mempool to allocate 'struct dm_io'
objects from.  As such, it saves an extra mempool allocation for each
original bio that DM core is issued.

This complicates the per-bio-data accessor functions by needing to
conditonally add extra padding to get to a target's per-bio-data.  But
in the end this provides a decent performance improvement for all
bio-based DM devices.

On an NVMe-loop based testbed to a ramdisk (~3100 MB/s): bio-based
DM linear performance improved by 2% (went from 2665 to 2777 MB/s).

Signed-off-by: Mike Snitzer

dm: remove unused 'num_write_bios' target interface

2017-12-13T17:15:58+00:00

No DM target provides num_write_bios and none has since dm-cache's
brief use in 2013.

Having the possibility of num_write_bios > 1 complicates bio
allocation.  So remove the interface and assume there is only one bio
needed.

If a target ever needs more, it must provide a suitable bioset and
allocate itself based on its particular needs.

Signed-off-by: NeilBrown 
Signed-off-by: Mike Snitzer