MD_RECOVERY_DONE is normally cleared by md_check_recovery after a
resync etc finished.  However it is possible for raid5_start_reshape
to race and start a reshape before MD_RECOVERY_DONE is cleared.  This
can lean to multiple reshapes running at the same time, which isn't
good.

To make sure it is cleared before starting a reshape, and also clear
it when reaping a thread, just to be safe.

Signed-off-by: NeilBrown  <neilb@suse.de>

Checking ->sync_thread without holding the mddev_lock()
isn't really safe, even after flushing the workqueue which
ensures md_start_sync() has been run.

While this code is waiting for the lock, md_check_recovery could reap
the thread itself, and then start another thread (e.g. recovery might
finish, then reshape starts).  When this thread gets the lock
md_start_sync() hasn't run so it doesn't get reaped, but
MD_RECOVERY_RUNNING gets cleared.  This allows two threads to start
which leads to confusion.

So don't both if MD_RECOVERY_RUNNING isn't set, but if it is do
the flush and the test and the reap all under the mddev_lock to
avoid any race with md_check_recovery.

Signed-off-by: NeilBrown <neilb@suse.de>
Fixes: 6791875e2e53 ("md: make reconfig_mutex optional for writes to md sysfs files.")
Cc: stable@vger.kernel.org (v4.0+)

Returning zero from a 'store' function is bad.
The return value should be either len length of the string
or an error.

So use 'len' if 'err' is zero.

Fixes: 6791875e2e53 ("md: make reconfig_mutex optional for writes to md sysfs files.")
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@vger.kernel (v4.0+)

Pull m,ore md bugfixes gfrom Neil Brown:
 "Assorted fixes for new RAID5 stripe-batching functionality.

  Unfortunately this functionality was merged a little prematurely.  The
  necessary testing and code review is now complete (or as complete as
  it can be) and to code passes a variety of tests and looks quite
  sensible.

  Also a fix for some recent locking changes - a race was introduced
  which causes a reshape request to sometimes fail.  No data safety
  issues"

* tag 'md/4.1-rc5-fixes' of git://neil.brown.name/md:
  md: fix race when unfreezing sync_action
  md/raid5: break stripe-batches when the array has failed.
  md/raid5: call break_stripe_batch_list from handle_stripe_clean_event
  md/raid5: be more selective about distributing flags across batch.
  md/raid5: add handle_flags arg to break_stripe_batch_list.
  md/raid5: duplicate some more handle_stripe_clean_event code in break_stripe_batch_list
  md/raid5: remove condition test from check_break_stripe_batch_list.
  md/raid5: Ensure a batch member is not handled prematurely.
  md/raid5: close race between STRIPE_BIT_DELAY and batching.
  md/raid5: ensure whole batch is delayed for all required bitmap updates.

A recent change removed the need for locking around writing
to "sync_action" (and various other places), but introduced a
subtle race.
When e.g. setting 'reshape' on a 'frozen' array, the 'frozen'
flag is cleared before 'reshape' is set, so the md thread can
get in and start trying recovery - which isn't wanted.

So instead of clearing MD_RECOVERY_FROZEN for any command
except 'frozen', only clear it when each specific command
is parsed.  This allows the handling of 'reshape' to clear
the bit while a lock is held.

Also remove some places where we set MD_RECOVERY_NEEDED,
as it is always set on non-error exit of the function.


Signed-off-by: NeilBrown <neilb@suse.de>
Fixes: 6791875e2e53 ("md: make reconfig_mutex optional for writes to md sysfs files.")

Pull block fixes from Jens Axboe:
 "A collection of fixes since the merge window;

   - fix for a double elevator module release, from Chao Yu.  Ancient bug.

   - the splice() MORE flag fix from Christophe Leroy.

   - a fix for NVMe, fixing a patch that went in in the merge window.
     From Keith.

   - two fixes for blk-mq CPU hotplug handling, from Ming Lei.

   - bdi vs blockdev lifetime fix from Neil Brown, fixing and oops in md.

   - two blk-mq fixes from Shaohua, fixing a race on queue stop and a
     bad merge issue with FUA writes.

   - division-by-zero fix for writeback from Tejun.

   - a block bounce page accounting fix, making sure we inc/dec after
     bouncing so that pre/post IO pages match up.  From Wang YanQing"

* 'for-linus' of git://git.kernel.dk/linux-block:
  splice: sendfile() at once fails for big files
  blk-mq: don't lose requests if a stopped queue restarts
  blk-mq: fix FUA request hang
  block: destroy bdi before blockdev is unregistered.
  block:bounce: fix call inc_|dec_zone_page_state on different pages confuse value of NR_BOUNCE
  elevator: fix double release of elevator module
  writeback: use |1 instead of +1 to protect against div by zero
  blk-mq: fix CPU hotplug handling
  blk-mq: fix race between timeout and CPU hotplug
  NVMe: Fix VPD B0 max sectors translation

Because of the peculiar way that md devices are created (automatically
when the device node is opened), a new device can be created and
registered immediately after the
	blk_unregister_region(disk_devt(disk), disk->minors);
call in del_gendisk().

Therefore it is important that all visible artifacts of the previous
device are removed before this call.  In particular, the 'bdi'.

Since:
commit c4db59d31e39ea067c32163ac961e9c80198fd37
Author: Christoph Hellwig <hch@lst.de>
    fs: don't reassign dirty inodes to default_backing_dev_info

moved the
   device_unregister(bdi->dev);
call from bdi_unregister() to bdi_destroy() it has been quite easy to
lose a race and have a new (e.g.) "md127" be created after the
blk_unregister_region() call and before bdi_destroy() is ultimately
called by the final 'put_disk', which must come after del_gendisk().

The new device finds that the bdi name is already registered in sysfs
and complains

> [ 9627.630029] WARNING: CPU: 18 PID: 3330 at fs/sysfs/dir.c:31 sysfs_warn_dup+0x5a/0x70()
> [ 9627.630032] sysfs: cannot create duplicate filename '/devices/virtual/bdi/9:127'

We can fix this by moving the bdi_destroy() call out of
blk_release_queue() (which can happen very late when a refcount
reaches zero) and into blk_cleanup_queue() - which happens exactly when the md
device driver calls it.

Then it is only necessary for md to call blk_cleanup_queue() before
del_gendisk().  As loop.c devices are also created on demand by
opening the device node, we make the same change there.

Fixes: c4db59d31e39ea067c32163ac961e9c80198fd37
Reported-by: Azat Khuzhin <a3at.mail@gmail.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org (v4.0)
Signed-off-by: NeilBrown <neilb@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>

When md notices non-sync IO happening while it is trying
to resync (or reshape or recover) it slows down to the
set minimum.

The default minimum might have made sense many years ago
but the drives have become faster.  Changing the default
to match the times isn't really a long term solution.

This patch changes the code so that instead of waiting until the speed
has dropped to the target, it just waits until pending requests
have completed.
This means that the delay inserted is a function of the speed
of the devices.

Testing shows that:
 - for some loads, the resync speed is unchanged.  For those loads
   increasing the minimum doesn't change the speed either.
   So this is a good result.  To increase resync speed under such
   loads we would probably need to increase the resync window
   size.

 - for other loads, resync speed does increase to a reasonable
   fraction (e.g. 20%) of maximum possible, and throughput of
   the load only drops a little bit (e.g. 10%)

 - for other loads, throughput of the non-sync load drops quite a bit
   more.  These seem to be latency-sensitive loads.

So it isn't a perfect solution, but it is mostly an improvement.

Signed-off-by: NeilBrown <neilb@suse.de>

This option is not well justified and testing suggests that
it hardly ever makes any difference.

The comment suggests there might be a need to wait for non-resync
activity indicated by ->nr_waiting, however raise_barrier()
already waits for all of that.

So just remove it to simplify reasoning about speed limiting.

This allows us to remove a 'FIXME' comment from raid5.c as that
never used the flag.

Signed-off-by: NeilBrown <neilb@suse.de>

There is really no need for sync_min to be a multiple of
chunk_size, and values read from here often aren't.
That means you cannot read a value and expect to be able
to write it back later.

So remove the chunk_size check, and round down to a multiple
of 4K, to be sure everything works with 4K-sector devices.

Signed-off-by: NeilBrown <neilb@suse.de>