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 c4db59d31e
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: c4db59d31e
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>
When "re-add" is writted to /sys/block/mdXX/md/dev-YYY/state,
the clustered md:
1. Sends RE_ADD message with the desc_nr. Nodes receiving the message
clear the Faulty bit in their respective rdev->flags.
2. The node initiating re-add, gathers the bitmaps of all nodes
and copies them into the local bitmap. It does not clear the bitmap
from which it is copying.
3. Initiating node schedules a md recovery to sync the devices.
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This adds the capability of re-adding a failed disk by
writing "re-add" to /sys/block/mdXX/md/dev-YYY/state.
This facilitates adding disks which have encountered a temporary
error such as a network disconnection/hiccup in an iSCSI device,
or a SAN cable disconnection which has been restored. In such
a situation, you do not need to remove and re-add the device.
Writing re-add to the failed device's state would add it again
to the array and perform the recovery of only the blocks which
were written after the device failed.
This works for generic md, and is not related to clustering. However,
this patch is to ease re-add operations listed above in clustering
environments.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This adds "remove" capabilities for the clustered environment.
When a user initiates removal of a device from the array, a
REMOVE message with disk number in the array is sent to all
the nodes which kick the respective device in their own array.
This facilitates the removal of failed devices.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This is required by the clustering module (patches to follow) to
find the device to remove or re-add.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This export is required for clustering module in order to
co-ordinate remove/readd a rdev from all nodes.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Simon reported the md io stats accounting issue:
"
I'm seeing "iostat -x -k 1" print this after a RAID1 rebuild on 4.0-rc5.
It's not abnormal other than it's 3-disk, with one being SSD (sdc) and
the other two being write-mostly:
Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util
sda 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
sdb 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
sdc 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
md0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 345.00 0.00 0.00 0.00 0.00 100.00
md2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 58779.00 0.00 0.00 0.00 0.00 100.00
md1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 12.00 0.00 0.00 0.00 0.00 100.00
"
The cause is commit "18c0b223cf9901727ef3b02da6711ac930b4e5d4" uses the
generic_start_io_acct to account the disk stats rather than the open code,
but it also introduced the increase to .in_flight[rw] which is needless to
md. So we re-use the open code here to fix it.
Reported-by: Simon Kirby <sim@hostway.ca>
Cc: <stable@vger.kernel.org> 3.19
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Signed-off-by: NeilBrown <neilb@suse.de>
A --cluster-confirm without an --add (by another node) can
crash the kernel.
Fix it by guarding it using a state.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
If ->run() fails, it can either free the data structures it
allocated, or leave that task to ->free() which will be called
on failures.
However:
md.c calls ->free() even if ->private_data is NULL, which
causes problems in some personalities.
raid0.c frees the data, but doesn't clear ->private_data,
which will become a problem when we fix md.c
So better fix both these issues at once.
Reported-by: Richard W.M. Jones <rjones@redhat.com>
Fixes: 5aa61f427e
URL: https://bugzilla.kernel.org/show_bug.cgi?id=94381
Signed-off-by: NeilBrown <neilb@suse.de>
Recent change to bitmap_create mishandles errors.
In particular a failure doesn't alway cause 'err' to be set.
Signed-off-by: NeilBrown <neilb@suse.de>
Since __ATTR_PREALLOC was introduced in v3.19-rc1~78^2~18
it can now be used by md.
This ensure that writing to these sysfs attributes will never
block due to a memory allocation.
Such blocking could become a deadlock if mdmon is trying to
reconfigure an array after a failure prior to re-enabling writes.
Signed-off-by: NeilBrown <neilb@suse.de>
Algorithm:
1. Node 1 issues mdadm --manage /dev/mdX --add /dev/sdYY which issues
ioctl(ADD_NEW_DISC with disc.state set to MD_DISK_CLUSTER_ADD)
2. Node 1 sends NEWDISK with uuid and slot number
3. Other nodes issue kobject_uevent_env with uuid and slot number
(Steps 4,5 could be a udev rule)
4. In userspace, the node searches for the disk, perhaps
using blkid -t SUB_UUID=""
5. Other nodes issue either of the following depending on whether the disk
was found:
ioctl(ADD_NEW_DISK with disc.state set to MD_DISK_CANDIDATE and
disc.number set to slot number)
ioctl(CLUSTERED_DISK_NACK)
6. Other nodes drop lock on no-new-devs (CR) if device is found
7. Node 1 attempts EX lock on no-new-devs
8. If node 1 gets the lock, it sends METADATA_UPDATED after unmarking the disk
as SpareLocal
9. If not (get no-new-dev lock), it fails the operation and sends METADATA_UPDATED
10. Other nodes understand if the device is added or not by reading the superblock again after receiving the METADATA_UPDATED message.
Signed-off-by: Lidong Zhong <lzhong@suse.com>
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
If there is a resync going on, all nodes must suspend writes to the
range. This is recorded in the suspend_info/suspend_list.
If there is an I/O within the ranges of any of the suspend_info,
should_suspend will return 1.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
When a resync is initiated, RESYNCING message is sent to all active
nodes with the range (lo,hi). When the resync is over, a RESYNCING
message is sent with (0,0). A high sector value of zero indicates
that the resync is over.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Re-reads the devices by invalidating the cache.
Since we don't write to faulty devices, this is detected using
events recorded in the devices. If it is old as compared to the mddev
mark it is faulty.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
- request to send a message
- make changes to superblock
- send messages telling everyone that the superblock has changed
- other nodes all read the superblock
- other nodes all ack the messages
- updating node release the "I'm sending a message" resource.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
When a node joins, it does not know of other nodes performing resync.
So, each node keeps the resync information in it's LVB. When a new
node joins, it reads the LVB of each "online" bitmap.
[TODO] The new node attempts to get the PW lock on other bitmap, if
it is successful, it reads the bitmap and performs the resync (if
required) on it's behalf.
If the node does not get the PW, it requests CR and reads the LVB
for the resync information.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
DLM offers callbacks when a node fails and the lock remastery
is performed:
1. recover_prep: called when DLM discovers a node is down
2. recover_slot: called when DLM identifies the node and recovery
can start
3. recover_done: called when all nodes have completed recover_slot
recover_slot() and recover_done() are also called when the node joins
initially in order to inform the node with its slot number. These slot
numbers start from one, so we deduct one to make it start with zero
which the cluster-md code uses.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
md_cluster_info stores the cluster information in the MD device.
The join() is called when mddev detects it is a clustered device.
The main responsibilities are:
1. Setup a DLM lockspace
2. Setup all initial locks such as super block locks and bitmap lock (will come later)
The leave() clears up the lockspace and all the locks held.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Rather than using mddev_lock() to take the reconfig_mutex
when writing to any md sysfs file, we only take mddev_lock()
in the particular _store() functions that require it.
Admittedly this is most, but it isn't all.
This also allows us to remove special-case handling for new_dev_store
(in md_attr_store).
Signed-off-by: NeilBrown <neilb@suse.de>
The one which is not inline (mddev_unlock) gets EXPORTed.
This makes the locking available to personality modules so that it
doesn't have to be imposed upon them.
Signed-off-by: NeilBrown <neilb@suse.de>
There are interdependencies between these two sysfs attributes
and whether a resync is currently running.
Rather than depending on reconfig_mutex to ensure no races when
testing these interdependencies are met, use the spinlock.
This will allow the mutex to be remove from protecting this
code in a subsequent patch.
Signed-off-by: NeilBrown <neilb@suse.de>
There isn't really much room for races with ->safemode_delay.
But as I am trying to clean up any racy code and will soon
be removing reconfig_mutex protection from most _store()
functions:
- only set mddev->safemode_delay once, to ensure no code
can see an intermediate value
- use safemode_timer to call md_safemode_timeout() rather than
calling it directly, to ensure it never races with itself.
Signed-off-by: NeilBrown <neilb@suse.de>
It makes more sense to report bitmap_info->file, rather than
bitmap->file (the later is only available once the array is
active).
With that change, use mddev->lock to protect bitmap_info being
set to NULL, and we can call get_bitmap_file() without taking
the mutex.
Signed-off-by: NeilBrown <neilb@suse.de>
1/ delay setting mddev->bitmap_info.file until 'f' looks
usable, so we don't have to unset it.
2/ Don't allow bitmap file to be set if bitmap_info.file
is already set.
Signed-off-by: NeilBrown <neilb@suse.de>
'buf' is only used because d_path fills from the end of the
buffer instead of from the start.
We don't need a separate buf to handle that, we just need to use
memmove() to move the string to the start.
Signed-off-by: NeilBrown <neilb@suse.de>
No rdev attributes need locking for 'show', though
state_show() might benefit from ensuring it sees a
consistent set of flags.
None even use rdev->mddev, so testing for it isn't really
needed and it certainly doesn't need to be held constant.
So improve state_show() and remove the locking.
Signed-off-by: NeilBrown <neilb@suse.de>
Most attributes can be read safely without any locking.
A race might lead to a slightly out-dated value, but nothing wrong.
We already have locking in some places where needed.
All that remains is can_clear_show(), behind_writes_used_show()
and action_show() which are easily fixed.
Signed-off-by: NeilBrown <neilb@suse.de>
The only access in md_seq_show that could suffer from races
not protected by ->lock is walking the rdev list.
This can receive sufficient protection from 'rcu'.
So use rdev_for_each_rcu() and get rid of mddev_lock().
Now reading /proc/mdstat will never block in md_seq_show.
Signed-off-by: NeilBrown <neilb@suse.de>
->pers is already protected by ->reconfig_mutex, and
cannot possibly change when there are threads running or
outstanding IO.
However there are some places where we access ->pers
not in a thread or IO context, and where ->reconfig_mutex
is unnecessarily heavy-weight: level_show and md_seq_show().
So protect all changes, and those accesses, with ->lock.
This is a step toward taking those accesses out from under
reconfig_mutex.
[Fixed missing "mddev->pers" -> "pers" conversion, thanks to
Dan Carpenter <dan.carpenter@oracle.com>]
Signed-off-by: NeilBrown <neilb@suse.de>
Gather all the changes that can happen atomically and might
be relevant to other code into one place. This will
make it easier to refine the locking.
Note that this puts quite a few things between mddev_detach()
and ->free(). Enabling this was the point of some recent patches.
Signed-off-by: NeilBrown <neilb@suse.de>
Now that the ->stop function only frees the private data,
rename is accordingly.
Also pass in the private pointer as an arg rather than using
mddev->private. This flexibility will be useful in level_store().
Finally, don't clear ->private. It doesn't make sense to clear
it seeing that isn't what we free, and it is no longer necessary
to clear ->private (it was some time ago before ->to_remove was
introduced).
Setting ->to_remove in ->free() is a bit of a wart, but not a
big problem at the moment.
Signed-off-by: NeilBrown <neilb@suse.de>
Each md personality has a 'stop' operation which does two
things:
1/ it finalizes some aspects of the array to ensure nothing
is accessing the ->private data
2/ it frees the ->private data.
All the steps in '1' can apply to all arrays and so can be
performed in common code.
This is useful as in the case where we change the personality which
manages an array (in level_store()), it would be helpful to do
step 1 early, and step 2 later.
So split the 'step 1' functionality out into a new mddev_detach().
Signed-off-by: NeilBrown <neilb@suse.de>
There is no locking around calls to merge_bvec_fn(), so
it is possible that calls which coincide with a level (or personality)
change could go wrong.
So create a central dispatch point for these functions and use
rcu_read_lock().
If the array is suspended, reject any merge that can be rejected.
If not, we know it is safe to call the function.
Signed-off-by: NeilBrown <neilb@suse.de>
There is currently no locking around calls to the 'congested'
bdi function. If called at an awkward time while an array is
being converted from one level (or personality) to another, there
is a tiny chance of running code in an unreferenced module etc.
So add a 'congested' function to the md_personality operations
structure, and call it with appropriate locking from a central
'mddev_congested'.
When the array personality is changing the array will be 'suspended'
so no IO is processed.
If mddev_congested detects this, it simply reports that the
array is congested, which is a safe guess.
As mddev_suspend calls synchronize_rcu(), mddev_congested can
avoid races by included the whole call inside an rcu_read_lock()
region.
This require that the congested functions for all subordinate devices
can be run under rcu_lock. Fortunately this is the case.
Signed-off-by: NeilBrown <neilb@suse.de>
This lock is used for (slightly) more than helping with writing
superblocks, and it will soon be extended further. So the
name is inappropriate.
Also, the _irq variant hasn't been needed since 2.6.37 as it is
never taking from interrupt or bh context.
So:
-rename write_lock to lock
-document what it protects
-remove _irq ... except in md_flush_request() as there
is no wait_event_lock() (with no _irq). This can be
cleaned up after appropriate changes to wait.h.
Signed-off-by: NeilBrown <neilb@suse.de>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2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=Ar2y
-----END PGP SIGNATURE-----
Merge tag 'md/3.19' of git://neil.brown.name/md
Pull md updates from Neil Brown:
"Three fixes for md.
I did have a largish set of locking changes queued, but late testing
showed they weren't quite as stable as I thought and while I fixed
what I found, I decided it safer to delay them a release ...
particularly as I'll be AFK for a few weeks. So expect a larger batch
next time :-)"
* tag 'md/3.19' of git://neil.brown.name/md:
md: Check MD_RECOVERY_RUNNING as well as ->sync_thread.
md: fix semicolon.cocci warnings
md/raid5: fetch_block must fetch all the blocks handle_stripe_dirtying wants.
A recent change to md started the ->sync_thread from a asynchronously
from a work_queue rather than synchronously. This means that there
can be a small window between the time when MD_RECOVERY_RUNNING is set
and when ->sync_thread is set.
So code that checks ->sync_thread might now conclude that the thread
has not been started and (because a lock is held) will not be started.
That is no longer the case.
Most of those places are best fixed by testing MD_RECOVERY_RUNNING
as well. To make this completely reliable, we wake_up(&resync_wait)
after clearing that flag as well as after clearing ->sync_thread.
Other places are better served by flushing the relevant workqueue
to ensure that that if the sync thread was starting, it has now
started. This is particularly best if we are about to stop the
sync thread.
Fixes: ac05f25669
Signed-off-by: NeilBrown <neilb@suse.de>
md_check_recovery will skip any recovery and also clear
MD_RECOVERY_NEEDED if MD_RECOVERY_FROZEN is set.
So when we clear _FROZEN, we must set _NEEDED and ensure that
md_check_recovery gets run.
Otherwise we could miss out on something that is needed.
In particular, this can make it impossible to remove a
failed device from an array is the 'recovery-needed' processing
didn't happen.
Suitable for stable kernels since 3.13.
Cc: stable@vger.kernel.org (3.13+)
Reported-and-tested-by: Joe Lawrence <joe.lawrence@stratus.com>
Fixes: 30b8feb730
Signed-off-by: NeilBrown <neilb@suse.de>
Linus Torvalds