'target_request_nr' is a more generic name that reflects the fact that
it will be used for both flush and discard support.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This change unifies the various checks and finalization that occurs on a
table prior to use. By doing so, it allows table construction without
traversing the dm-ioctl interface.
Signed-off-by: Will Drewry <wad@chromium.org>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Implement merge method for the snapshot origin to improve read
performance.
Without merge method, dm asks the upper layers to submit smallest possible
bios --- one page. Submitting such small bios impacts performance negatively
when reading or writing the origin device.
Without this patch, CPU consumption when reading the origin on lvm on md-raid0
was 6 to 12%, with this patch, it drops to 1 to 4%.
Note: in my testing, it actually degraded performance in some settings, I
traced it to Maxtor disks having problems with > 512-sector requests.
Reducing the number of sectors to /sys/block/sd*/queue/max_sectors_kb to
256 fixed the read performance. I think we don't have to care about weird
disks that actually degrade performance because of large requests being
sent to them.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Change bio-based mapped devices no longer to have a fully initialized
request_queue (request_fn, elevator, etc). This means bio-based DM
devices no longer register elevator sysfs attributes ('iosched/' tree
or 'scheduler' other than "none").
In contrast, a request-based DM device will continue to have a full
request_queue and will register elevator sysfs attributes. Therefore
a user can determine a DM device's type by checking if elevator sysfs
attributes exist.
First allocate a minimalist request_queue structure for a DM device
(needed for both bio and request-based DM).
Initialization of a full request_queue is deferred until it is known
that the DM device is request-based, at the end of the table load
sequence.
Factor DM device's request_queue initialization:
- common to both request-based and bio-based into dm_init_md_queue().
- specific to request-based into dm_init_request_based_queue().
The md->type_lock mutex is used to protect md->queue, in addition to
md->type, during table_load().
A DM device's first table_load will establish the immutable md->type.
But md->queue initialization, based on md->type, may fail at that time
(because blk_init_allocated_queue cannot allocate memory). Therefore
any subsequent table_load must (re)try dm_setup_md_queue independently of
establishing md->type.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Determine whether a mapped device is bio-based or request-based when
loading its first (inactive) table and don't allow that to be changed
later.
This patch performs different device initialisation in each of the two
cases. (We don't think it's necessary to add code to support changing
between the two types.)
Allowed md->type transitions:
DM_TYPE_NONE to DM_TYPE_BIO_BASED
DM_TYPE_NONE to DM_TYPE_REQUEST_BASED
We now prevent table_load from replacing the inactive table with a
conflicting type of table even after an explicit table_clear.
Introduce 'type_lock' into the struct mapped_device to protect md->type
and to prepare for the next patch that will change the queue
initialization and allocate memory while md->type_lock is held.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
drivers/md/dm-ioctl.c | 15 +++++++++++++++
drivers/md/dm.c | 37 ++++++++++++++++++++++++++++++-------
drivers/md/dm.h | 5 +++++
include/linux/dm-ioctl.h | 4 ++--
4 files changed, 52 insertions(+), 9 deletions(-)
When processing barriers, skip the second flush if processing the bio
failed with -EOPNOTSUPP. This can happen with discard+barrier requests.
If the device doesn't support discard, there would be two useless
SYNCHRONIZE CACHE commands. The first dm_flush cannot be so easily
optimized out, so we leave it there.
Previously, -EOPNOTSUPP could be received in dec_pending only with empty
barriers and we ignored that error, assuming the device not supporting
cache flushes has cache always consistent. With the addition of discard
barriers, this -EOPNOTSUPP can also be generated by discards and we
must record it in md->barrier_error for process_barrier.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch fixes hard-coded value for the size of a chunk that includes
disk header for persistent snapshot. It should be changed to existing
macro NUM_SNAPSHOT_HDR_CHUNKS instead of using hard-coded value 1.
Signed-off-by: Tomohiro Kusumi <kusumi.tomohiro@jp.fujitsu.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Use kstrdup when the goal of an allocation is copy a string into the
allocated region.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@
expression from,to;
expression flag,E1,E2;
statement S;
@@
- to = kmalloc(strlen(from) + 1,flag);
+ to = kstrdup(from, flag);
... when != \(from = E1 \| to = E1 \)
if (to==NULL || ...) S
... when != \(from = E2 \| to = E2 \)
- strcpy(to, from);
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
The dm control device does not implement read/write, so it has no use for
seeking. Using no_llseek prevents falling back to default_llseek, which
requires the BKL.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch separates the device deletion code from dm_put()
to make sure the deletion happens in the process context.
By this patch, device deletion always occurs in an ioctl (process)
context and dm_put() can be called in interrupt context.
As a result, the request-based dm's bad dm_put() usage pointed out
by Mikulas below disappears.
http://marc.info/?l=dm-devel&m=126699981019735&w=2
Without this patch, I confirmed there is a case to crash the system:
dm_put() => dm_table_destroy() => vfree() => BUG_ON(in_interrupt())
Some more backgrounds and details:
In request-based dm, a device opener can remove a mapped_device
while the last request is still completing, because bios in the last
request complete first and then the device opener can close and remove
the mapped_device before the last request completes:
CPU0 CPU1
=================================================================
<<INTERRUPT>>
blk_end_request_all(clone_rq)
blk_update_request(clone_rq)
bio_endio(clone_bio) == end_clone_bio
blk_update_request(orig_rq)
bio_endio(orig_bio)
<<I/O completed>>
dm_blk_close()
dev_remove()
dm_put(md)
<<Free md>>
blk_finish_request(clone_rq)
....
dm_end_request(clone_rq)
free_rq_clone(clone_rq)
blk_end_request_all(orig_rq)
rq_completed(md)
So request-based dm used dm_get()/dm_put() to hold md for each I/O
until its request completion handling is fully done.
However, the final dm_put() can call the device deletion code which
must not be run in interrupt context and may cause kernel panic.
To solve the problem, this patch moves the device deletion code,
dm_destroy(), to predetermined places that is actually deleting
the mapped_device in ioctl (process) context, and changes dm_put()
just to decrement the reference count of the mapped_device.
By this change, dm_put() can be used in any context and the symmetric
model below is introduced:
dm_create(): create a mapped_device
dm_destroy(): destroy a mapped_device
dm_get(): increment the reference count of a mapped_device
dm_put(): decrement the reference count of a mapped_device
dm_destroy() waits for all references of the mapped_device to disappear,
then deletes the mapped_device.
dm_destroy() uses active waiting with msleep(1), since deleting
the mapped_device isn't performance-critical task.
And since at this point, nobody opens the mapped_device and no new
reference will be taken, the pending counts are just for racing
completing activity and will eventually decrease to zero.
For the unlikely case of the forced module unload, dm_destroy_immediate(),
which doesn't wait and forcibly deletes the mapped_device, is also
introduced and used in dm_hash_remove_all(). Otherwise, "rmmod -f"
may be stuck and never return.
And now, because the mapped_device is deleted at this point, subsequent
accesses to the mapped_device may cause NULL pointer references.
Cc: stable@kernel.org
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch changes dm_hash_remove_all() to release _hash_lock when
removing a device. After removing the device, dm_hash_remove_all()
takes _hash_lock and searches the hash from scratch again.
This patch is a preparation for the next patch, which changes device
deletion code to wait for md reference to be 0. Without this patch,
the wait in the next patch may cause AB-BA deadlock:
CPU0 CPU1
-----------------------------------------------------------------------
dm_hash_remove_all()
down_write(_hash_lock)
table_status()
md = find_device()
dm_get(md)
<increment md->holders>
dm_get_live_or_inactive_table()
dm_get_inactive_table()
down_write(_hash_lock)
<in the md deletion code>
<wait for md->holders to be 0>
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch prevents access to mapped_device which is being deleted.
Currently, even after a mapped_device has been removed from the hash,
it could be accessed through idr_find() using minor number.
That could cause a race and NULL pointer reference below:
CPU0 CPU1
------------------------------------------------------------------
dev_remove(param)
down_write(_hash_lock)
dm_lock_for_deletion(md)
spin_lock(_minor_lock)
set_bit(DMF_DELETING)
spin_unlock(_minor_lock)
__hash_remove(hc)
up_write(_hash_lock)
dev_status(param)
md = find_device(param)
down_read(_hash_lock)
__find_device_hash_cell(param)
dm_get_md(param->dev)
md = dm_find_md(dev)
spin_lock(_minor_lock)
md = idr_find(MINOR(dev))
spin_unlock(_minor_lock)
dm_put(md)
free_dev(md)
dm_get(md)
up_read(_hash_lock)
__dev_status(md, param)
dm_put(md)
This patch fixes such problems.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
All the dm ioctls that generate uevents set the DM_UEVENT_GENERATED flag so
that userspace knows whether or not to wait for a uevent to be processed
before continuing,
The dm rename ioctl sets this flag but was not structured to return it
to userspace. This patch restructures the rename ioctl processing to
behave like the other ioctls that return data and so fix this.
Signed-off-by: Peter Rajnoha <prajnoha@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Remove useless __dev_status call while processing an ioctl that sets up
device geometry and target message. The data is not returned to
userspace so there is no point collecting it and in the case of
target_message it is collected before processing the message so if it
did return it might be stale.
Signed-off-by: Peter Rajnoha <prajnoha@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Validate chunk size against both origin and snapshot sector size
Don't allow chunk size smaller than either origin or snapshot logical
sector size. Reading or writing data not aligned to sector size is not
allowed and causes immediate errors.
This requires us to open the origin before initialising the
exception store and to export dm_snap_origin.
Cc: stable@kernel.org
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Iterate both origin and snapshot devices
iterate_devices method should call the callback for all the devices where
the bio may be remapped. Thus, snapshot_iterate_devices should call the callback
for both snapshot and origin underlying devices because it remaps some bios
to the snapshot and some to the origin.
snapshot_iterate_devices called the callback only for the origin device.
This led to badly calculated device limits if snapshot and origin were placed
on different types of disks.
Cc: stable@kernel.org
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
multipath_ctr() forgets to return an error after detecting
missing path parameters. Fix this.
Signed-off-by: Patrick LoPresti <lopresti@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
* 'for-linus' of git://neil.brown.name/md: (24 commits)
md: clean up do_md_stop
md: fix another deadlock with removing sysfs attributes.
md: move revalidate_disk() back outside open_mutex
md/raid10: fix deadlock with unaligned read during resync
md/bitmap: separate out loading a bitmap from initialising the structures.
md/bitmap: prepare for storing write-intent-bitmap via dm-dirty-log.
md/bitmap: optimise scanning of empty bitmaps.
md/bitmap: clean up plugging calls.
md/bitmap: reduce dependence on sysfs.
md/bitmap: white space clean up and similar.
md/raid5: export raid5 unplugging interface.
md/plug: optionally use plugger to unplug an array during resync/recovery.
md/raid5: add simple plugging infrastructure.
md/raid5: export is_congested test
raid5: Don't set read-ahead when there is no queue
md: add support for raising dm events.
md: export various start/stop interfaces
md: split out md_rdev_init
md: be more careful setting MD_CHANGE_CLEAN
md/raid5: ensure we create a unique name for kmem_cache when mddev has no gendisk
...
There is only one error exit from do_md_stop, so make that more
explicit and discard the 'err' variable.
Also drop the 'revalidate' variable by moving the unlock calls around.
Signed-off-by: NeilBrown <neilb@suse.de>
Move the deletion of sysfs attributes from reconfig_mutex to
open_mutex didn't really help as a process can try to take
open_mutex while holding reconfig_mutex, so the same deadlock can
happen, just requiring one more process to be involved in the chain.
I looks like I cannot easily use locking to wait for the sysfs
deletion to complete, so don't.
The only things that we cannot do while the deletions are still
pending is other things which can change the sysfs namespace: run,
takeover, stop. Each of these can fail with -EBUSY.
So set a flag while doing a sysfs deletion, and fail run, takeover,
stop if that flag is set.
This is suitable for 2.6.35.x
Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
Commit b821eaa5 "md: remove ->changed and related code" moved
revalidate_disk() under open_mutex, and lockdep noticed.
[ INFO: possible circular locking dependency detected ]
2.6.32-mdadm-locking #1
-------------------------------------------------------
mdadm/3640 is trying to acquire lock:
(&bdev->bd_mutex){+.+.+.}, at: [<ffffffff811acecb>] revalidate_disk+0x5b/0x90
but task is already holding lock:
(&mddev->open_mutex){+.+...}, at: [<ffffffffa055e07a>] do_md_stop+0x4a/0x4d0 [md_mod]
which lock already depends on the new lock.
It is suitable for 2.6.35.x
Cc: <stable@kernel.org>
Reported-by: Przemyslaw Czarnowski <przemyslaw.hawrylewicz.czarnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
The open and release block_device_operations are currently
called with the BKL held. In order to change that, we must
first make sure that all drivers that currently rely
on this have no regressions.
This blindly pushes the BKL into all .open and .release
operations for all block drivers to prepare for the
next step. The drivers can subsequently replace the BKL
with their own locks or remove it completely when it can
be shown that it is not needed.
The functions blkdev_get and blkdev_put are the only
remaining users of the big kernel lock in the block
layer, besides a few uses in the ioctl code, none
of which need to serialize with blkdev_{get,put}.
Most of these two functions is also under the protection
of bdev->bd_mutex, including the actual calls to
->open and ->release, and the common code does not
access any global data structures that need the BKL.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Remove the current bio flags and reuse the request flags for the bio, too.
This allows to more easily trace the type of I/O from the filesystem
down to the block driver. There were two flags in the bio that were
missing in the requests: BIO_RW_UNPLUG and BIO_RW_AHEAD. Also I've
renamed two request flags that had a superflous RW in them.
Note that the flags are in bio.h despite having the REQ_ name - as
blkdev.h includes bio.h that is the only way to go for now.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Remove all the trivial wrappers for the cmd_type and cmd_flags fields in
struct requests. This allows much easier grepping for different request
types instead of unwinding through macros.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
If the 'bio_split' path in raid10-read is used while
resync/recovery is happening it is possible to deadlock.
Fix this be elevating ->nr_waiting for the duration of both
parts of the split request.
This fixes a bug that has been present since 2.6.22
but has only started manifesting recently for unknown reasons.
It is suitable for and -stable since then.
Reported-by: Justin Bronder <jsbronder@gentoo.org>
Tested-by: Justin Bronder <jsbronder@gentoo.org>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
dm makes this distinction between ->ctr and ->resume, so we need to
too.
Also get the new bitmap_load to clear out the bitmap first, as this is
most consistent with the dm suspend/resume approach
Signed-off-by: NeilBrown <neilb@suse.de>
This allows md/raid5 to fully work as a dm target.
Normally md uses a 'filemap' which contains a list of pages of bits
each of which may be written separately.
dm-log uses and all-or-nothing approach to writing the log, so
when using a dm-log, ->filemap is NULL and the flags normally stored
in filemap_attr are stored in ->logattrs instead.
Signed-off-by: NeilBrown <neilb@suse.de>
A bitmap is stored as one page per 2048 bits.
If none of the bits are set, the page is not allocated.
When bitmap_get_counter finds that a page isn't allocate,
it just reports that one bit work of space isn't flagged,
rather than reporting that 2048 bits worth of space are
unflagged.
This can cause searches for flagged bits (e.g. bitmap_close_sync)
to do more work than is really necessary.
So change bitmap_get_counter (when creating) to report a number of
blocks that more accurately reports the range of the device for which
no counter currently exists.
Signed-off-by: NeilBrown <neilb@suse.de>
1/ use md_unplug in bitmap.c as we will soon be using bitmaps under
arrays with no queue attached.
2/ Don't bother plugging the queue when we set a bit in the bitmap.
The reason for this was to encourage as many bits as possible to
get set before we unplug and write stuff out.
However every personality already plugs the queue after
bitmap_startwrite either directly (raid1/raid10) or be setting
STRIPE_BIT_DELAY which causes the queue to be plugged later
(raid5).
Signed-off-by: NeilBrown <neilb@suse.de>
For dm-raid45 we will want to use bitmaps in dm-targets which don't
have entries in sysfs, so cope with the mddev not living in sysfs.
Signed-off-by: NeilBrown <neilb@suse.de>
Fixes some whitespace problems
Fixed some checkpatch.pl complaints.
Replaced kmalloc ... memset(0), with kzalloc
Fixed an unlikely memory leak on an error path.
Reformatted a number of 'if/else' sets, sometimes
replacing goto with an else clause.
Removed some old comments and commented-out code.
Signed-off-by: NeilBrown <neilb@suse.de>
If an array doesn't have a 'queue' then md_do_sync cannot
unplug it.
In that case it will have a 'plugger', so make that available
to the mddev, and use it to unplug the array if needed.
Signed-off-by: NeilBrown <neilb@suse.de>
md/raid5 uses the plugging infrastructure provided by the block layer
and 'struct request_queue'. However when we plug raid5 under dm there
is no request queue so we cannot use that.
So create a similar infrastructure that is much lighter weight and use
it for raid5.
Signed-off-by: NeilBrown <neilb@suse.de>
the dm module will need this for dm-raid45.
Also only access ->queue->backing_dev_info->congested_fn
if ->queue actually exists. It won't in a dm target.
Signed-off-by: NeilBrown <neilb@suse.de>
dm-raid456 does not provide a 'queue' for raid5 to use,
so we must make raid5 stop depending on the queue.
First: read_ahead
dm handles read-ahead adjustment fully in userspace, so
simply don't do any readahead adjustments if there is
no queue.
Also re-arrange code slightly so all the accesses to ->queue are
together.
Finally, move the blk_queue_merge_bvec function into the 'if' as
the ->split_io setting in dm-raid456 has the same effect.
Signed-off-by: NeilBrown <neilb@suse.de>
dm uses scheduled work to raise events to user-space.
So allow md device to have work_structs and schedule them on an error.
Signed-off-by: NeilBrown <neilb@suse.de>
export entry points for starting and stopping md arrays.
This will be used by a module to make md/raid5 work under
dm.
Also stop calling md_stop_writes from md_stop, as that won't
work well with dm - it will want to call the two separately.
Signed-off-by: NeilBrown <neilb@suse.de>
This functionality will be needed separately in a subsequent patch, so
split it into it's own exported function.
Signed-off-by: NeilBrown <neilb@suse.de>
When MD_CHANGE_CLEAN is set we might block in md_write_start.
So we should only set it when fairly sure that something will clear
it.
There are two places where it is set so as to encourage a metadata
update to record the progress of resync/recovery. This should only
be done if the internal metadata update mechanisms are in use, which
can be tested by by inspecting '->persistent'.
Signed-off-by: NeilBrown <neilb@suse.de>
We will shortly allow md devices with no gendisk (they are attached to
a dm-target instead). That will cause mdname() to return 'mdX'.
There is one place where mdname really needs to be unique: when
creating the name for a slab cache.
So in that case, if there is no gendisk, you the address of the mddev
formatted in HEX to provide a unique name.
Signed-off-by: NeilBrown <neilb@suse.de>
We will want md devices to live as dm targets where sysfs is not
visible. So allow md to not connect to sysfs.
Signed-off-by: NeilBrown <neilb@suse.de>
There are few situations where it would make any sense to add a spare
when reducing the number of devices in an array, but it is
conceivable: A 6 drive RAID6 with two missing devices could be
reshaped to a 5 drive RAID6, and a spare could become available
just in time for the reshape, but not early enough to have been
recovered first. 'freezing' recovery can make this easy to
do without any races.
However doing such a thing is a bad idea. md will not record the
partially-recovered state of the 'spare' and when the reshape
finished it will think that the spare is still spare.
Easiest way to avoid this confusion is to simply disallow it.
Signed-off-by: NeilBrown <neilb@suse.de>
As the comment says, the tail of this loop only applies to devices
that are not fully in sync, so if In_sync was set, we should avoid
the rest of the loop.
This bug will hardly ever cause an actual problem. The worst it
can do is allow an array to be assembled that is dirty and degraded,
which is not generally a good idea (without warning the sysadmin
first).
This will only happen if the array is RAID4 or a RAID5/6 in an
intermediate state during a reshape and so has one drive that is
all 'parity' - no data - while some other device has failed.
This is certainly possible, but not at all common.
Signed-off-by: NeilBrown <neilb@suse.de>
During a recovery of reshape the early part of some devices might be
in-sync while the later parts are not.
We we know we are looking at an early part it is good to treat that
part as in-sync for stripe calculations.
This is particularly important for a reshape which suffers device
failure. Treating the data as in-sync can mean the difference between
data-safety and data-loss.
Signed-off-by: NeilBrown <neilb@suse.de>
When we are reshaping an array, the device failure combinations
that cause us to decide that the array as failed are more subtle.
In particular, any 'spare' will be fully in-sync in the section
of the array that has already been reshaped, thus failures that
affect only that section are less critical.
So encode this subtlety in a new function and call it as appropriate.
The case that showed this problem was a 4 drive RAID5 to 8 drive RAID6
conversion where the last two devices failed.
This resulted in:
good good good good incomplete good good failed failed
while converting a 5-drive RAID6 to 8 drive RAID5
The incomplete device causes the whole array to look bad,
bad as it was actually good for the section that had been
converted to 8-drives, all the data was actually safe.
Reported-by: Terry Morris <tbmorris@tbmorris.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When an array is reshaped to have fewer devices, the reshape proceeds
from the end of the devices to the beginning.
If a device happens to be non-In_sync (which is possible but rare)
we would normally update the ->recovery_offset as the reshape
progresses. However that would be wrong as the recover_offset records
that the early part of the device is in_sync, while in fact it would
only be the later part that is in_sync, and in any case the offset
number would be measured from the wrong end of the device.
Relatedly, if after a reshape a spare is discovered to not be
recoverred all the way to the end, not allow spare_active
to incorporate it in the array.
This becomes relevant in the following sample scenario:
A 4 drive RAID5 is converted to a 6 drive RAID6 in a combined
operation.
The RAID5->RAID6 conversion will cause a 5 drive to be included as a
spare, then the 5drive -> 6drive reshape will effectively rebuild that
spare as it progresses. The 6th drive is treated as in_sync the whole
time as there is never any case that we might consider reading from
it, but must not because there is no valid data.
If we interrupt this reshape part-way through and reverse it to return
to a 5-drive RAID6 (or event a 4-drive RAID5), we don't want to update
the recovery_offset - as that would be wrong - and we don't want to
include that spare as active in the 5-drive RAID6 when the reversed
reshape completed and it will be mostly out-of-sync still.
Signed-off-by: NeilBrown <neilb@suse.de>
The entries in the stripe_cache maintained by raid5 are enlarged
when we increased the number of devices in the array, but not
shrunk when we reduce the number of devices.
So if entries are added after reducing the number of devices, we
much ensure to initialise the whole entry, not just the part that
is currently relevant. Otherwise if we enlarge the array again,
we will reference uninitialised values.
As grow_buffers/shrink_buffer now want to use a count that is stored
explicity in the raid_conf, they should get it from there rather than
being passed it as a parameter.
Signed-off-by: NeilBrown <neilb@suse.de>
Only level 5 with layout=PARITY_N can be taken over to raid0 now.
Lets allow level 4 either.
Signed-off-by: Maciej Trela <maciej.trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
After takeover from raid5/10 -> raid0 mddev->layout is not cleared.
Signed-off-by: Maciej Trela <maciej.trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Use mddev->new_layout in setup_conf.
Also use new_chunk, and don't set ->degraded in takeover(). That
gets set in run()
Signed-off-by: Maciej Trela <maciej.trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Most array level changes leave the list of devices largely unchanged,
possibly causing one at the end to become redundant.
However conversions between RAID0 and RAID10 need to renumber
all devices (except 0).
This renumbering is currently being done in the ->run method when the
new personality takes over. However this is too late as the common
code in md.c might already have invalidated some of the devices if
they had a ->raid_disk number that appeared to high.
Moving it into the ->takeover method is too early as the array is
still active at that time and wrong ->raid_disk numbers could cause
confusion.
So add a ->new_raid_disk field to mdk_rdev_s and use it to communicate
the new raid_disk number.
Now the common code knows exactly which devices need to be renumbered,
and which can be invalidated, and can do it all at a convenient time
when the array is suspend.
It can also update some symlinks in sysfs which previously were not be
updated correctly.
Reported-by: Maciej Trela <maciej.trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Such NULL pointer dereference can occur when the driver was fixing the
read errors/bad blocks and the disk was physically removed
causing a system crash. This patch check if the
rcu_dereference() returns valid rdev before accessing it in fix_read_error().
Cc: stable@kernel.org
Signed-off-by: Prasanna S. Panchamukhi <prasanna.panchamukhi@riverbed.com>
Signed-off-by: Rob Becker <rbecker@riverbed.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Commit b821eaa572 broke partition
detection for md arrays.
The logic was almost right. However if revalidate_disk is called
when the device is not yet open, bdev->bd_disk won't be set, so the
flush_disk() Call will not set bd_invalidated.
So when md_open is called we still need to ensure that
->bd_invalidated gets set. This is easily done with a call to
check_disk_size_change in the place where the offending commit removed
check_disk_change. At the important times, the size will have changed
from 0 to non-zero, so check_disk_size_change will set bd_invalidated.
Tested-by: Duncan <1i5t5.duncan@cox.net>
Reported-by: Duncan <1i5t5.duncan@cox.net>
Signed-off-by: NeilBrown <neilb@suse.de>
By the previous modification, the cpu notifier can return encapsulate
errno value. This converts the cpu notifiers for raid5.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (69 commits)
fix handling of offsets in cris eeprom.c, get rid of fake on-stack files
get rid of home-grown mutex in cris eeprom.c
switch ecryptfs_write() to struct inode *, kill on-stack fake files
switch ecryptfs_get_locked_page() to struct inode *
simplify access to ecryptfs inodes in ->readpage() and friends
AFS: Don't put struct file on the stack
Ban ecryptfs over ecryptfs
logfs: replace inode uid,gid,mode initialization with helper function
ufs: replace inode uid,gid,mode initialization with helper function
udf: replace inode uid,gid,mode init with helper
ubifs: replace inode uid,gid,mode initialization with helper function
sysv: replace inode uid,gid,mode initialization with helper function
reiserfs: replace inode uid,gid,mode initialization with helper function
ramfs: replace inode uid,gid,mode initialization with helper function
omfs: replace inode uid,gid,mode initialization with helper function
bfs: replace inode uid,gid,mode initialization with helper function
ocfs2: replace inode uid,gid,mode initialization with helper function
nilfs2: replace inode uid,gid,mode initialization with helper function
minix: replace inode uid,gid,mode init with helper
ext4: replace inode uid,gid,mode init with helper
...
Trivial conflict in fs/fs-writeback.c (mark bitfields unsigned)
Conflicts:
drivers/md/md.c
- Resolved conflict in md_update_sb
- Added extra 'NULL' arg to new instance of sysfs_get_dirent.
Signed-off-by: NeilBrown <neilb@suse.de>
Now that the last user passing a NULL file pointer is gone we can remove
the redundant dentry argument and associated hacks inside vfs_fsynmc_range.
The next step will be removig the dentry argument from ->fsync, but given
the luck with the last round of method prototype changes I'd rather
defer this until after the main merge window.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The problem. When implementing a network namespace I need to be able
to have multiple network devices with the same name. Currently this
is a problem for /sys/class/net/*, /sys/devices/virtual/net/*, and
potentially a few other directories of the form /sys/ ... /net/*.
What this patch does is to add an additional tag field to the
sysfs dirent structure. For directories that should show different
contents depending on the context such as /sys/class/net/, and
/sys/devices/virtual/net/ this tag field is used to specify the
context in which those directories should be visible. Effectively
this is the same as creating multiple distinct directories with
the same name but internally to sysfs the result is nicer.
I am calling the concept of a single directory that looks like multiple
directories all at the same path in the filesystem tagged directories.
For the networking namespace the set of directories whose contents I need
to filter with tags can depend on the presence or absence of hotplug
hardware or which modules are currently loaded. Which means I need
a simple race free way to setup those directories as tagged.
To achieve a reace free design all tagged directories are created
and managed by sysfs itself.
Users of this interface:
- define a type in the sysfs_tag_type enumeration.
- call sysfs_register_ns_types with the type and it's operations
- sysfs_exit_ns when an individual tag is no longer valid
- Implement mount_ns() which returns the ns of the calling process
so we can attach it to a sysfs superblock.
- Implement ktype.namespace() which returns the ns of a syfs kobject.
Everything else is left up to sysfs and the driver layer.
For the network namespace mount_ns and namespace() are essentially
one line functions, and look to remain that.
Tags are currently represented a const void * pointers as that is
both generic, prevides enough information for equality comparisons,
and is trivial to create for current users, as it is just the
existing namespace pointer.
The work needed in sysfs is more extensive. At each directory
or symlink creating I need to check if the directory it is being
created in is a tagged directory and if so generate the appropriate
tag to place on the sysfs_dirent. Likewise at each symlink or
directory removal I need to check if the sysfs directory it is
being removed from is a tagged directory and if so figure out
which tag goes along with the name I am deleting.
Currently only directories which hold kobjects, and
symlinks are supported. There is not enough information
in the current file attribute interfaces to give us anything
to discriminate on which makes it useless, and there are
no potential users which makes it an uninteresting problem
to solve.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Benjamin Thery <benjamin.thery@bull.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Devices which know that they are spares do not really need to have
an event count that matches the rest of the array, so there are no
data-in-sync issues. It is enough that the uuid matches.
So remove the requirement that the event count is up-to-date.
We currently still write out and event count on spares, but this
allows us in a year or 3 to stop doing that completely.
Signed-off-by: NeilBrown <neilb@suse.de>
When updating the event count for a simple clean <-> dirty transition,
we try to avoid updating the spares so they can safely spin-down.
As the event_counts across an array must be +/- 1, this means
decrementing the event_count on a dirty->clean transition.
This is not always safe and we have to avoid the unsafe time.
We current do this with a misguided idea about it being safe or
not depending on whether the event_count is odd or even. This
approach only works reliably in a few common instances, but easily
falls down.
So instead, simply keep internal state concerning whether it is safe
or not, and always assume it is not safe when an array is first
assembled.
Signed-off-by: NeilBrown <neilb@suse.de>
Fix: Raid-6 was not trying to correct a read-error when in
singly-degraded state and was instead dropping one more device, going to
doubly-degraded state. This patch fixes this behaviour.
Tested-by: Janos Haar <janos.haar@netcenter.hu>
Signed-off-by: Gabriele A. Trombetti <g.trombetti.lkrnl1213@logicschema.com>
Reported-by: Janos Haar <janos.haar@netcenter.hu>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Some time ago we stopped the clean/active metadata updates
from being written to a 'spare' device in most cases so that
it could spin down and say spun down. Device failure/removal
etc are still recorded on spares.
However commit 51d5668cb2 broke this 50% of the time,
depending on whether the event count is even or odd.
The change log entry said:
This means that the alignment between 'odd/even' and
'clean/dirty' might take a little longer to attain,
how ever the code makes no attempt to create that alignment, so it
could take arbitrarily long.
So when we find that clean/dirty is not aligned with odd/even,
force a second metadata-update immediately. There are already cases
where a second metadata-update is needed immediately (e.g. when a
device fails during the metadata update). We just piggy-back on that.
Reported-by: Joe Bryant <tenminjoe@yahoo.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
read_balance uses a "unsigned long" for a sector number which
will get truncated beyond 2TB.
This will cause read-balancing to be non-optimal, and can cause
data to be read from the 'wrong' branch during a resync. This has a
very small chance of returning wrong data.
Reported-by: Jordan Russell <jr-list-2010@quo.to>
Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
Many 'printk' messages from the raid456 module mention 'raid5' even
though it may be a 'raid6' or even 'raid4' array. This can cause
confusion.
Also the actual array name is not always reported and when it is
it is not reported consistently.
So change all the messages to start:
md/raid:%s:
where '%s' becomes e.g. md3 to identify the particular array.
Signed-off-by: NeilBrown <neilb@suse.de>
RAID10 has been available for quite a while now and is quite well
tested, so we can remove the EXPERIMENTAL designation.
Reported-by: Eric MSP Veith <eveith@wwweb-library.net>
Signed-off-by: NeilBrown <neilb@suse.de>
Level modifications change the output of mdstat. The mdmon manager
thread is interested in these events for external metadata management.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
When a raid1 array is configured to support write-behind
on some devices, it normally only reads from other devices.
If all devices are write-behind (because the rest have failed)
it is possible for a read request to be serviced before a
behind-write request, which would appear as data corruption.
So when forced to read from a WriteMostly device, wait for any
write-behind to complete, and don't start any more behind-writes.
Signed-off-by: NeilBrown <neilb@suse.de>
This message seems to suggest the named device is the one on which a
read failed, however it is actually the device that the read will be
redirected to.
So make the message a little clearer.
Reported-by: Tim Burgess <ozburgess@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
This is
- unnecessary because mddev_suspend is always followed by a call to
->stop, and each ->stop unregisters the thread, and
- a problem as it makes it awkwards to suspend and then resume a
device as we will want later.
Signed-off-by: NeilBrown <neilb@suse.de>
We used to pass the personality make_request function direct
to the block layer so the first argument had to be a queue.
But now we have the intermediary md_make_request so it makes
at lot more sense to pass a struct mddev_s.
It makes it possible to have an mddev without its own queue too.
Signed-off-by: NeilBrown <neilb@suse.de>
This moves the call to the other side of set_readonly, but that should
not be an issue.
This encapsulates in 'md_stop' all of the functionality for internally
stopping the array, leaving all the interactions with externalities
(sysfs, request_queue, gendisk) in do_md_stop.
Signed-off-by: NeilBrown <neilb@suse.de>
Using do_md_stop to set an array to read-only is a little confusing.
Now most of the common code has been factored out, split
md_set_readonly off in to a separate function.
Signed-off-by: NeilBrown <neilb@suse.de>
Further refactoring of do_md_stop.
This one requires some explanation as it takes code from different
places in do_md_stop, so some re-ordering happens.
We only get into this part of do_md_stop if there are no active opens
of the device, so no writes can be happening and the device must have
been flushed. In md_stop_writes we want to stop any internal sources
of writes - i.e. resync - and flush out the metadata.
The only code that was previously before some of this code is
code to clean up the queue, the mddev, the gendisk, or sysfs, all
of which is probably better after code that makes active changes (i.e.
triggers writes).
Signed-off-by: NeilBrown <neilb@suse.de>
do_md_stop is large and clunky, so hard to understand.
This is a first step of refactoring, pulling two simple
sub-functions out.
Signed-off-by: NeilBrown <neilb@suse.de>
As part of relaxing the binding between an mddev and gendisk,
we separate do_md_run into two functions.
md_run does all the work internal to md
do_md_run calls md_run and makes and changes to gendisk
that are required.
Signed-off-by: NeilBrown <neilb@suse.de>
We set ->changed to 1 and call check_disk_change at the end
of md_open so that bd_invalidated would be set and thus
partition rescan would happen appropriately.
Now that we call revalidate_disk directly, which sets bd_invalidates,
that indirection is no longer needed and can be removed.
Signed-off-by: NeilBrown <neilb@suse.de>
Using ->array_sectors rather than get_capacity() is more
direct and is a step towards relaxing the tight connection
between mddev and gendisk.
Signed-off-by: NeilBrown <neilb@suse.de>
While I generally prefer letting personalities do as much as possible,
given that we have a central md_make_request anyway we may as well use
it to simplify code.
Also this centralises knowledge of ->gendisk which will help later.
Signed-off-by: NeilBrown <neilb@suse.de>
Diving through ->queue to find mddev is unnecessarily complex - there
is an easier path to finding mddev, so use that.
Signed-off-by: NeilBrown <neilb@suse.de>
Level changes can be very significant, so make sure
to notify them via sysfs.
Signed-off-by: Maciej Trela <maciej.trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When metadata is being managed by user-space, md doesn't know
what the maximum number of devices allowed in an array is
so ->max_disks is 0. In this case we should allow any (+ve)
number of disks.
Signed-off-by: NeilBrown <neilb@suse.de>
Writing "none" to "../md/dev-xx/slot" removes that device
from being an active part of the array, but it didn't
set ->raid_disk to -1 to record this fact.
Signed-off-by: Maciej Trela <Maciej.Trela@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
In a subsequent patch we will make it possible to change
mddev->raid_disks while a RAID0 or RAID10 array is active. This is
part of the process of reshaping such an array.
This means that we cannot use this value while processes requests
(it is OK to use it during initialisation as we are locked against
changes then).
Both RAID0 and RAID10 have the same value stored in the private data
structure, so use that value instead.
Signed-off-by: NeilBrown <neilb@suse.de>
This was needed when sysfs files could only be 'notified'
from process context. Now that we have sys_notify_direct,
we can call it directly from an interrupt.
Signed-off-by: NeilBrown <neilb@suse.de>
void pointers do not need to be cast to other pointer types.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Keep track of the maximum number of concurrent write-behind requests
for an md array and exposed this number in sysfs at
md/bitmap/max_backlog_used
Writing any value to this file will clear it.
This allows userspace to be involved in tuning bitmap/backlog.
Signed-off-by: Paul Clements <paul.clements@steeleye.com>
Signed-off-by: NeilBrown <neilb@suse.de>
These fields have never been used.
commit 4b6d287f62
added them, but also added identical files to bitmap_super_s,
and only used the latter.
So remove these unused fields.
Signed-off-by: NeilBrown <neilb@suse.de>
There is a very small race window when writing to a
RAID1 such that if a device is marked faulty at exactly the wrong
time, the write-in-progress will not be sent to the device,
but the bitmap (if present) will be updated to say that
the write was sent.
Then if the device turned out to still be usable as was re-added
to the array, the bitmap-based-resync would skip resyncing that
block, possibly leading to corruption. This would only be a problem
if no further writes were issued to that area of the device (i.e.
that bitmap chunk).
Suitable for any pending -stable kernel.
Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
Some levels expect the 'redundancy group' to be present,
others don't.
So when we change level of an array we might need to
add or remove this group.
This requires fixing up the current practice of overloading ->private
to indicate (when ->pers == NULL) that something needs to be removed.
So create a new ->to_remove to fill that role.
When changing levels, we may need to add or remove attributes. When
changing RAID5 -> RAID6, we both add and remove the same thing. It is
important to catch this and optimise it out as the removal is delayed
until a lock is released, so trying to add immediately would cause
problems.
Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
When an array is stopped we need to remove some
sysfs files which are dependent on the type of array.
We need to delay that deletion as deleting them while holding
reconfig_mutex can lead to deadlocks.
We currently delay them until the array is completely destroyed.
However it is possible to deactivate and then reactivate the array.
It is also possible to need to remove sysfs files when changing level,
which can potentially happen several times before an array is
destroyed.
So we need to delete these files more promptly: as soon as
reconfig_mutex is dropped.
We need to ensure this happens before do_md_run can restart the array,
so we use open_mutex for some extra locking. This is not deadlock
prone.
Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
Since commit ef286f6fa6
it has been important that each personality clears
->private in the ->stop() function, or sets it to a
attribute group to be removed.
linear.c doesn't. This can sometimes lead to an oops,
though it doesn't always.
Suitable for 2.6.33-stable and 2.6.34.
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
When the user sets the block device to readwrite then the mddev should
follow suit. Otherwise, the BUG_ON in md_write_start() will be set to
trigger.
The reverse direction, setting mddev->ro to match a set readonly
request, can be ignored because the blkdev level readonly flag precludes
the need to have mddev->ro set correctly. Nevermind the fact that
setting mddev->ro to 1 may fail if the array is in use.
Cc: <stable@kernel.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Some time ago we stopped the clean/active metadata updates
from being written to a 'spare' device in most cases so that
it could spin down and say spun down. Device failure/removal
etc are still recorded on spares.
However commit 51d5668cb2 broke this 50% of the time,
depending on whether the event count is even or odd.
The change log entry said:
This means that the alignment between 'odd/even' and
'clean/dirty' might take a little longer to attain,
how ever the code makes no attempt to create that alignment, so it
could take arbitrarily long.
So when we find that clean/dirty is not aligned with odd/even,
force a second metadata-update immediately. There are already cases
where a second metadata-update is needed immediately (e.g. when a
device fails during the metadata update). We just piggy-back on that.
Reported-by: Joe Bryant <tenminjoe@yahoo.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Fix: Raid-6 was not trying to correct a read-error when in
singly-degraded state and was instead dropping one more device, going to
doubly-degraded state. This patch fixes this behaviour.
Tested-by: Janos Haar <janos.haar@netcenter.hu>
Signed-off-by: Gabriele A. Trombetti <g.trombetti.lkrnl1213@logicschema.com>
Reported-by: Janos Haar <janos.haar@netcenter.hu>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Previous patch changes stripe and chunk_number to sector_t but
mistakenly did not update all of the divisions to use sector_dev().
This patch changes all the those divisions (actually the '%' operator)
to sector_div.
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Tested-by: Stefan Lippers-Hollmann <s.l-h@gmx.de>
With many large drives and small chunk sizes it is possible
to create a RAID5 with more than 2^31 chunks. Make sure this
works.
Reported-by: Brett King <king.br@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
If a component device has a merge_bvec_fn then as we never call it
we must ensure we never need to. Currently this is done by setting
max_sector to 1 PAGE, however this does not stop a bio being created
with several sub-page iovecs that would violate the merge_bvec_fn.
So instead set max_segments to 1 and set the segment boundary to the
same as a page boundary to ensure there is only ever one single-page
segment of IO requested at a time.
This can particularly be an issue when 'xen' is used as it is
known to submit multiple small buffers in a single bio.
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Constify struct sysfs_ops.
This is part of the ops structure constification
effort started by Arjan van de Ven et al.
Benefits of this constification:
* prevents modification of data that is shared
(referenced) by many other structure instances
at runtime
* detects/prevents accidental (but not intentional)
modification attempts on archs that enforce
read-only kernel data at runtime
* potentially better optimized code as the compiler
can assume that the const data cannot be changed
* the compiler/linker move const data into .rodata
and therefore exclude them from false sharing
Signed-off-by: Emese Revfy <re.emese@gmail.com>
Acked-by: David Teigland <teigland@redhat.com>
Acked-by: Matt Domsch <Matt_Domsch@dell.com>
Acked-by: Maciej Sosnowski <maciej.sosnowski@intel.com>
Acked-by: Hans J. Koch <hjk@linutronix.de>
Acked-by: Pekka Enberg <penberg@cs.helsinki.fi>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
To prevent deadlock, bios in the hold list should be flushed before
dm_rh_stop_recovery() is called in mirror_suspend().
The recovery can't start because there are pending bios and therefore
dm_rh_stop_recovery deadlocks.
When there are pending bios in the hold list, the recovery waits for
the completion of the bios after recovery_count is acquired.
The recovery_count is released when the recovery finished, however,
the bios in the hold list are processed after dm_rh_stop_recovery() in
mirror_presuspend(). dm_rh_stop_recovery() also acquires recovery_count,
then deadlock occurs.
Signed-off-by: Takahiro Yasui <tyasui@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Eliminate a 4-byte hole in 'struct dm_io_memory' by moving 'offset' above the
'ptr' to which it applies (size reduced from 24 to 16 bytes). And by
association, 1-4 byte hole is eliminated in 'struct dm_io_request' (size
reduced from 56 to 48 bytes).
Eliminate all 6 4-byte holes and 1 cache-line in 'struct dm_snapshot' (size
reduced from 392 to 368 bytes).
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Set a new DM_UEVENT_GENERATED_FLAG when returning from ioctls to
indicate that a uevent was actually generated. This tells the userspace
caller that it may need to wait for the event to be processed.
Signed-off-by: Peter Rajnoha <prajnoha@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Free the dm_io structure before calling bio_endio() instead of after it,
to ensure that the io_pool containing it is not referenced after it is
freed.
This partially fixes a problem described here
https://www.redhat.com/archives/dm-devel/2010-February/msg00109.html
thread 1:
bio_endio(bio, io_error);
/* scheduling happens */
thread 2:
close the device
remove the device
thread 1:
free_io(md, io);
Thread 2, when removing the device, sees non-empty md->io_pool (because the
io hasn't been freed by thread 1 yet) and may crash with BUG in mempool_free.
Thread 1 may also crash, when freeing into a nonexisting mempool.
To fix this we must make sure that bio_endio() is the last call and
the md structure is not accessed afterwards.
There is another bio_endio in process_barrier, but it is called from the thread
and the thread is destroyed prior to freeing the mempools, so this call is
not affected by the bug.
A similar bug exists with module unloads - the module may be unloaded
immediately after bio_endio - but that is more difficult to fix.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Remove unused parameters(start and len) of dm_get_device()
and fix the callers.
Signed-off-by: Nikanth Karthikesan <knikanth@suse.de>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Only issue a uevent on a resume if the state of the device changed,
i.e. if it was suspended and/or its table was replaced.
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
If all mirror legs fail, always return an error instead of holding the
bio, even if the handle_errors option was set. At present it is the
responsibility of the driver underneath us to deal with retries,
multipath etc.
The patch adds the bio to the failures list instead of holding it
directly. do_failures tests first if all legs failed and, if so,
returns the bio with -EIO. If any leg is still alive and handle_errors
is set, do_failures calls hold_bio.
Reviewed-by: Takahiro Yasui <tyasui@redhat.com>
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch pulls the pg_init path activation code out of
process_queued_ios() into a new function.
No functional change.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
When suspending the device we must wait for all I/O to complete, but
pg-init may be still in progress even after flushing the workqueue
for kmpath_handlerd in multipath_postsuspend.
This patch waits for pg-init completion correctly in
multipath_postsuspend().
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
m->queue_io is set to block processing I/Os, and it needs to be kept
while pg-init, which issues multiple path activations, is in progress.
But m->queue is cleared when a path activation completes without error
in pg_init_done(), even while other path activations are in progress.
That may cause undesired -EIO on paths which are not complete activation.
This patch fixes that by not clearing m->queue_io until all path
activations complete.
(Before the hardware handlers were moved into the SCSI layer, pg_init
only used one path.)
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
'suspended' flag in struct multipath was introduced to check whether
the multipath target is in suspended state, but the same check is
done through dm_suspended() now, so remove the flag and related code.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Remove the dm_get() in dm_table_get_md() because dm_table_get_md() could
be called from presuspend/postsuspend, which are called while
mapped_device is in DMF_FREEING state, where dm_get() is not allowed.
Justification for that is the lifetime of both objects: As far as the
current dm design/implementation, mapped_device is never freed while
targets are doing something, because dm core waits for targets to become
quiet in dm_put() using presuspend/postsuspend. So targets should be
able to touch mapped_device without holding reference count of the
mapped_device, and we should allow targets to touch mapped_device even
if it is in DMF_FREEING state.
Backgrounds:
I'm trying to remove the multipath internal queue, since dm core now has
a generic queue for request-based dm. In the patch-set, the multipath
target wants to request dm core to start/stop queue. One of such
start/stop requests can happen during postsuspend() while the target
waits for pg-init to complete, because the target stops queue when
starting pg-init and tries to restart it when completing pg-init. Since
queue belongs to mapped_device, it involves calling dm_table_get_md()
and dm_put(). On the other hand, postsuspend() is called in dm_put()
for mapped_device which is in DMF_FREEING state, and that triggers
BUG_ON(DMF_FREEING) in the 2nd dm_put().
I had tried to solve this problem by changing only multipath not to
touch mapped_device which is in DMF_FREEING state, but I couldn't and I
came up with a question why we need dm_get() in dm_table_get_md().
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch adds two minor fixes while processing device mapper path activation.
Skip failed paths while calling activate_path. If the path is already failed
then activate_path will fail for sure. We don't have to call in that case. In
some case this might cause prolonged retries unnecessarily.
Change the misleading message if the path being activated fails with SCSI_DH_NOSYS.
Signed-off-by: Babu Moger <babu.moger@lsi.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch removes some unnecessary argument casting. There is no
functional change with this patch.
Passes 'struct pgpath' through to pg_init_done() instead of the enclosed
'struct dm_path'.
Tested the changes with LSI storage..
CC: Chandra Seetharaman <chandra.seetharaman@us.ibm.com>
Signed-off-by: Babu Moger <babu.moger@lsi.com>
Acked-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Except for SCSI no device drivers distinguish between physical and
hardware segment limits. Consolidate the two into a single segment
limit.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
The block layer calling convention is blk_queue_<limit name>.
blk_queue_max_sectors predates this practice, leading to some confusion.
Rename the function to appropriately reflect that its intended use is to
set max_hw_sectors.
Also introduce a temporary wrapper for backwards compability. This can
be removed after the merge window is closed.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Add __percpu sparse annotations to places which didn't make it in one
of the previous patches. All converions are trivial.
These annotations are to make sparse consider percpu variables to be
in a different address space and warn if accessed without going
through percpu accessors. This patch doesn't affect normal builds.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Borislav Petkov <borislav.petkov@amd.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Huang Ying <ying.huang@intel.com>
Cc: Len Brown <lenb@kernel.org>
Cc: Neil Brown <neilb@suse.de>
Revert commit d2bb7df8ca at Greg's request.
Author: Milan Broz <mbroz@redhat.com>
Date: Thu Dec 10 23:51:53 2009 +0000
dm: sysfs add empty release function to avoid debug warning
This patch just removes an unnecessary warning:
kobject: 'dm': does not have a release() function,
it is broken and must be fixed.
The kobject is embedded in mapped device struct, so
code does not need to release memory explicitly here.
Cc: Greg KH <gregkh@suse.de>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch fixes the problem that system may stall if target's ->map_rq
returns DM_MAPIO_REQUEUE in map_request().
E.g. stall happens on 1 CPU box when a dm-mpath device with queue_if_no_path
bounces between all-paths-down and paths-up on I/O load.
When target's ->map_rq returns DM_MAPIO_REQUEUE, map_request() requeues
the request and returns to dm_request_fn(). Then, dm_request_fn()
doesn't exit the I/O dispatching loop and continues processing
the requeued request again.
This map and requeue loop can be done with interrupt disabled,
so 1 CPU system can be stalled if this situation happens.
For example, commands below can stall my 1 CPU box within 1 minute or so:
# dmsetup table mp
mp: 0 2097152 multipath 1 queue_if_no_path 0 1 1 service-time 0 1 2 8:144 1 1
# while true; do dd if=/dev/mapper/mp of=/dev/null bs=1M count=100; done &
# while true; do \
> dmsetup message mp 0 "fail_path 8:144" \
> dmsetup suspend --noflush mp \
> dmsetup resume mp \
> dmsetup message mp 0 "reinstate_path 8:144" \
> done
To fix the problem above, this patch changes dm_request_fn() to exit
the I/O dispatching loop once if a request is requeued in map_request().
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
When suspending a failed mirror, bios are completed by mirror_end_io() and
__rh_lookup() in dm_rh_dec() returns NULL where a non-NULL return value is
required by design. Fix this by not changing the state of the recovery failed
region from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end().
Issue
On 2.6.33-rc1 kernel, I hit the bug when I suspended the failed
mirror by dmsetup command.
BUG: unable to handle kernel NULL pointer dereference at 00000020
IP: [<f94f38e2>] dm_rh_dec+0x35/0xa1 [dm_region_hash]
...
EIP: 0060:[<f94f38e2>] EFLAGS: 00010046 CPU: 0
EIP is at dm_rh_dec+0x35/0xa1 [dm_region_hash]
EAX: 00000286 EBX: 00000000 ECX: 00000286 EDX: 00000000
ESI: eff79eac EDI: eff79e80 EBP: f6915cd4 ESP: f6915cc4
DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process dmsetup (pid: 2849, ti=f6914000 task=eff03e80 task.ti=f6914000)
...
Call Trace:
[<f9530af6>] ? mirror_end_io+0x53/0x1b1 [dm_mirror]
[<f9413104>] ? clone_endio+0x4d/0xa2 [dm_mod]
[<f9530aa3>] ? mirror_end_io+0x0/0x1b1 [dm_mirror]
[<f94130b7>] ? clone_endio+0x0/0xa2 [dm_mod]
[<c02d6bcb>] ? bio_endio+0x28/0x2b
[<f952f303>] ? hold_bio+0x2d/0x62 [dm_mirror]
[<f952f942>] ? mirror_presuspend+0xeb/0xf7 [dm_mirror]
[<c02aa3e2>] ? vmap_page_range+0xb/0xd
[<f9414c8d>] ? suspend_targets+0x2d/0x3b [dm_mod]
[<f9414ca9>] ? dm_table_presuspend_targets+0xe/0x10 [dm_mod]
[<f941456f>] ? dm_suspend+0x4d/0x150 [dm_mod]
[<f941767d>] ? dev_suspend+0x55/0x18a [dm_mod]
[<c0343762>] ? _copy_from_user+0x42/0x56
[<f9417fb0>] ? dm_ctl_ioctl+0x22c/0x281 [dm_mod]
[<f9417628>] ? dev_suspend+0x0/0x18a [dm_mod]
[<f9417d84>] ? dm_ctl_ioctl+0x0/0x281 [dm_mod]
[<c02c3c4b>] ? vfs_ioctl+0x22/0x85
[<c02c422c>] ? do_vfs_ioctl+0x4cb/0x516
[<c02c42b7>] ? sys_ioctl+0x40/0x5a
[<c0202858>] ? sysenter_do_call+0x12/0x28
Analysis
When recovery process of a region failed, dm_rh_recovery_end() function
changes the state of the region from RM_RH_RECOVERING to DM_RH_NOSYNC.
When recovery_complete() is executed between dm_rh_update_states() and
dm_writes() in do_mirror(), bios are processed with the region state,
DM_RH_NOSYNC. However, the region data is freed without checking its
pending count when dm_rh_update_states() is called next time.
When bios are finished by mirror_end_io(), __rh_lookup() in dm_rh_dec()
returns NULL even though a valid return value are expected.
Solution
Remove the state change of the recovery failed region from DM_RH_RECOVERING
to DM_RH_NOSYNC in dm_rh_recovery_end(). We can remove the state change
because:
- If the region data has been released by dm_rh_update_states(),
a new region data is created with the state of DM_RH_NOSYNC, and
bios are processed according to the DM_RH_NOSYNC state.
- If the region data has not been released by dm_rh_update_states(),
a state of the region is DM_RH_RECOVERING and bios are put in the
delayed_bio list.
The flag change from DM_RH_RECOVERING to DM_RH_NOSYNC in dm_rh_recovery_end()
was added in the following commit:
dm raid1: handle resync failures
author Jonathan Brassow <jbrassow@redhat.com>
Thu, 12 Jul 2007 16:29:04 +0000 (17:29 +0100)
http://git.kernel.org/linus/f44db678edcc6f4c2779ac43f63f0b9dfa28b724
Signed-off-by: Takahiro Yasui <tyasui@redhat.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
If the mirror log fails when the handle_errors option was not selected
and there is no remaining valid mirror leg, writes return success even
though they weren't actually written to any device. This patch
completes them with EIO instead.
This code path is taken:
do_writes:
bio_list_merge(&ms->failures, &sync);
do_failures:
if (!get_valid_mirror(ms)) (false)
else if (errors_handled(ms)) (false)
else bio_endio(bio, 0);
The logic in do_failures is based on presuming that the write was already
tried: if it succeeded at least on one leg (without handle_errors) it
is reported as success.
Reference: https://bugzilla.redhat.com/show_bug.cgi?id=555197
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch fixes two bugs that revolve around the miscalculation and
misuse of the variable 'overhead_size'. 'overhead_size' is the size of
the various header structures used during communication.
The first bug is the use of 'sizeof' with the pointer of a structure
instead of the structure itself - resulting in the wrong size being
computed. This is then used in a check to see if the payload
(data_size) would be to large for the preallocated structure. Since the
bug produces a smaller value for the overhead, it was possible for the
structure to be breached. (Although the current users of the code do
not currently send enough data to trigger this bug.)
The second bug is that the 'overhead_size' value is used to compute how
much of the preallocated space should be cleared before populating it
with fresh data. This should have simply been 'sizeof(struct cn_msg)'
not overhead_size. The fact that 'overhead_size' was computed
incorrectly made this problem "less bad" - leaving only a pointer's
worth of space at the end uncleared. Thus, this bug was never producing
a bad result, but still needs to be fixed - especially now that the
value is computed correctly.
Cc: stable@kernel.org
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
chunk_io() declares its 'struct mdata_req' on the stack and then
initializes its 'struct work_struct' member. Annotate the
initialization of this workqueue with INIT_WORK_ON_STACK to suppress a
debugobjects warning seen when CONFIG_DEBUG_OBJECTS_WORK is enabled.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
If a table containing zero as stripe count is passed into stripe_ctr
the code attempts to divide by zero.
This patch changes DM_TABLE_LOAD to return -EINVAL if the stripe count
is zero.
We now get the following error messages:
device-mapper: table: 253:0: striped: Invalid stripe count
device-mapper: ioctl: error adding target to table
Signed-off-by: Nikanth Karthikesan <knikanth@suse.de>
Cc: stable@kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
======
This fix is related to
http://bugzilla.kernel.org/show_bug.cgi?id=15142
but does not address that exact issue.
======
sysfs does like attributes being removed while they are being accessed
(i.e. read or written) and waits for the access to complete.
As accessing some md attributes takes the same lock that is held while
removing those attributes a deadlock can occur.
This patch addresses 3 issues in md that could lead to this deadlock.
Two relate to calling flush_scheduled_work while the lock is held.
This is probably a bad idea in general and as we use schedule_work to
delete various sysfs objects it is particularly bad.
In one case flush_scheduled_work is called from md_alloc (called by
md_probe) called from do_md_run which holds the lock. This call is
only present to ensure that ->gendisk is set. However we can be sure
that gendisk is always set (though possibly we couldn't when that code
was originally written. This is because do_md_run is called in three
different contexts:
1/ from md_ioctl. This requires that md_open has succeeded, and it
fails if ->gendisk is not set.
2/ from writing a sysfs attribute. This can only happen if the
mddev has been registered in sysfs which happens in md_alloc
after ->gendisk has been set.
3/ from autorun_array which is only called by autorun_devices, which
checks for ->gendisk to be set before calling autorun_array.
So the call to md_probe in do_md_run can be removed, and the check on
->gendisk can also go.
In the other case flush_scheduled_work is being called in do_md_stop,
purportedly to wait for all md_delayed_delete calls (which delete the
component rdevs) to complete. However there really isn't any need to
wait for them - they have already been disconnected in all important
ways.
The third issue is that raid5->stop() removes some attribute names
while the lock is held. There is already some infrastructure in place
to delay attribute removal until after the lock is released (using
schedule_work). So extend that infrastructure to remove the
raid5_attrs_group.
This does not address all lockdep issues related to the sysfs
"s_active" lock. The rest can be address by splitting that lockdep
context between symlinks and non-symlinks which hopefully will happen.
Signed-off-by: NeilBrown <neilb@suse.de>
This code was written long ago when it was not possible to
reshape a degraded array. Now it is so the current level of
degraded-ness needs to be taken in to account. Also newly addded
devices should only reduce degradedness if they are deemed to be
in-sync.
In particular, if you convert a RAID5 to a RAID6, and increase the
number of devices at the same time, then the 5->6 conversion will
make the array degraded so the current code will produce a wrong
value for 'degraded' - "-1" to be precise.
If the reshape runs to completion end_reshape will calculate a correct
new value for 'degraded', but if a device fails during the reshape an
incorrect decision might be made based on the incorrect value of
"degraded".
This patch is suitable for 2.6.32-stable and if they are still open,
2.6.31-stable and 2.6.30-stable as well.
Cc: stable@kernel.org
Reported-by: Michael Evans <mjevans1983@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Make DM use bdev_stack_limits() function so that partition offsets get
taken into account when calculating alignment. Clarify stacking
warnings.
Also remove obsolete clearing of final alignment_offset and misalignment
flag.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: Alasdair G. Kergon <agk@redhat.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
If two arrays share a device, then they will not both resync at the
same time. One will wait for the other to complete.
While waiting, the MD_RECOVERY_INTR flag is not checked so a device
failure, which would make the resync pointless, does not cause the
resync to abort, so the failed device cannot be removed (as it cannot
be remove while a resync is happening).
So add a test for MD_RECOVERY_INTR.
Reported-by: Brett Russ <bruss@netezza.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Since commit dfc7064500,
->hot_remove_disks has not removed non-failed devices from
an array until recovery is no longer possible.
So the code in do_md_run to get around the fact that
md_check_recovery (which calls ->hot_remove_disks) would
remove partially-in-sync devices is no longer needed.
So remove it.
Signed-off-by: NeilBrown <neilb@suse.de>
By default md_do_sync() will perform recovery if no other actions are
specified. However, action_show() relies on MD_RECOVERY_RECOVER to be
set otherwise it returns 'idle'. So, add a missing set
MD_RECOVERY_RECOVER when starting recovery.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
The start_ro modules parameter can be used to force arrays to be
started in 'auto-readonly' in which they are read-only until the first
write. This ensures that no resync/recovery happens until something
else writes to the device. This is important for resume-from-disk
off an md array.
However if an array is started 'readonly' (by writing 'readonly' to
the 'array_state' sysfs attribute) we want it to be really 'readonly',
not 'auto-readonly'.
So strengthen the condition to only set auto-readonly if the
array is not already read-only.
Signed-off-by: NeilBrown <neilb@suse.de>
evms configures md arrays by:
open device
send ioctl
close device
for each different ioctl needed.
Since 2.6.29, the device can disappear after the 'close'
unless a significant configuration has happened to the device.
The change made by "SET_ARRAY_INFO" can too minor to stop the device
from disappearing, but important enough that losing the change is bad.
So: make sure SET_ARRAY_INFO sets mddev->ctime, and keep the device
active as long as ctime is non-zero (it gets zeroed with lots of other
things when the array is stopped).
This is suitable for -stable kernels since 2.6.29.
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
* git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-2.6-dm: (80 commits)
dm snapshot: use merge origin if snapshot invalid
dm snapshot: report merge failure in status
dm snapshot: merge consecutive chunks together
dm snapshot: trigger exceptions in remaining snapshots during merge
dm snapshot: delay merging a chunk until writes to it complete
dm snapshot: queue writes to chunks being merged
dm snapshot: add merging
dm snapshot: permit only one merge at once
dm snapshot: support barriers in snapshot merge target
dm snapshot: avoid allocating exceptions in merge
dm snapshot: rework writing to origin
dm snapshot: add merge target
dm exception store: add merge specific methods
dm snapshot: create function for chunk_is_tracked wait
dm snapshot: make bio optional in __origin_write
dm mpath: reject messages when device is suspended
dm: export suspended state to targets
dm: rename dm_suspended to dm_suspended_md
dm: swap target postsuspend call and setting suspended flag
dm crypt: add plain64 iv
...
Signed-off-by: Joe Perches <joe@perches.com>
Cc: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Makes use of skip_spaces() defined in lib/string.c for removing leading
spaces from strings all over the tree.
It decreases lib.a code size by 47 bytes and reuses the function tree-wide:
text data bss dec hex filename
64688 584 592 65864 10148 (TOTALS-BEFORE)
64641 584 592 65817 10119 (TOTALS-AFTER)
Also, while at it, if we see (*str && isspace(*str)), we can be sure to
remove the first condition (*str) as the second one (isspace(*str)) also
evaluates to 0 whenever *str == 0, making it redundant. In other words,
"a char equals zero is never a space".
Julia Lawall tried the semantic patch (http://coccinelle.lip6.fr) below,
and found occurrences of this pattern on 3 more files:
drivers/leds/led-class.c
drivers/leds/ledtrig-timer.c
drivers/video/output.c
@@
expression str;
@@
( // ignore skip_spaces cases
while (*str && isspace(*str)) { \(str++;\|++str;\) }
|
- *str &&
isspace(*str)
)
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Cc: Julia Lawall <julia@diku.dk>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jeff Dike <jdike@addtoit.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br>
Cc: David Howells <dhowells@redhat.com>
Cc: <linux-ext4@vger.kernel.org>
Cc: Samuel Ortiz <samuel@sortiz.org>
Cc: Patrick McHardy <kaber@trash.net>
Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Enable external metadata arrays to manage rebuild checkpointing via a
md/dev-XXX/recovery_start attribute which reflects rdev->recovery_offset
Also update resync_start_store to allow 'none' to be written, for
consistency.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Other walks of this list are either under rcu_read_lock() or the list
mutation lock (mddev_lock()). This protects against the improbable case of a
disk being removed from the array at the start of md_do_sync().
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
As v1.x metadata can record that a member of the array is
not completely recovered, it make sense to record that a
spare has become a regular member of the array at the earliest
opportunity.
So remove the tests on "recovery_offset > 0" in super_1_sync
as they really aren't needed, and schedule a metadata update
immediately after adding spares to a degraded array.
This means that if a crash happens immediately after a recovery
starts, the new device will be included in the array and recovery will
continue from wherever it was up to. Previously this didn't happen
unless recovery was at least 1/16 of the way through.
Signed-off-by: NeilBrown <neilb@suse.de>
The RAID ioctls are only implemented in md.c, so the
handling for them should also be moved there from
fs/compat_ioctl.c.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Neil Brown <neilb@suse.de>
Cc: Andre Noll <maan@systemlinux.org>
Cc: linux-raid@vger.kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>
Make it clear in the config message that MD_MULTIPATH is not under
active development.
Cc: Oren Held <orenhe@il.ibm.com>
Signed-off-by: NeilBrown <neilb@suse.de>
We've noticed severe lasting performance degradation of our raid
arrays when we have drives that yield large amounts of media errors.
The raid10 module will queue each failed read for retry, and also
will attempt call fix_read_error() to perform the read recovery.
Read recovery is performed while the array is frozen, so repeated
recovery attempts can degrade the performance of the array for
extended periods of time.
With this patch I propose adding a per md device max number of
corrected read attempts. Each rdev will maintain a count of
read correction attempts in the rdev->read_errors field (not
used currently for raid10). When we enter fix_read_error()
we'll check to see when the last read error occurred, and
divide the read error count by 2 for every hour since the
last read error. If at that point our read error count
exceeds the read error threshold, we'll fail the raid device.
In addition in this patch I add sysfs nodes (get/set) for
the per md max_read_errors attribute, the rdev->read_errors
attribute, and added some printk's to indicate when
fix_read_error fails to repair an rdev.
For testing I used debugfs->fail_make_request to inject
IO errors to the rdev while doing IO to the raid array.
Signed-off-by: Robert Becker <Rob.Becker@riverbed.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When we get a read error on a device in a RAID10, and attempting to
repair the error fails, print more useful messages about why it
failed.
Signed-off-by: Robert Becker <Rob.Becker@riverbed.com>
Signed-off-by: NeilBrown <neilb@suse.de>
There is a sysfs file which allows bits in the write-intent
bitmap to be explicit set - indicating that the block is thought
to be 'dirty'.
When this happens we should really set recovery_cp backwards
to include the block to reflect this dirtiness.
In particular, a 'resync' process will refuse to start if
recovery_cp is beyond the end of the array, so this is needed
to allow a resync to be triggered.
Signed-off-by: NeilBrown <neilb@suse.de>
In this case, the metadata needs to not be in the same
sector as the bitmap.
md will not read/write any bitmap metadata. Config must be
done via sysfs and when a recovery makes the array non-degraded
again, writing 'true' to 'bitmap/can_clear' will allow bits in
the bitmap to be cleared again.
Signed-off-by: NeilBrown <neilb@suse.de>
Setting daemon_lastrun really has nothing to do with reading
the bitmap superblock, it just happens to be needed at the same time.
bitmap_read_sb is about to become options, so move that code out
to after the call to bitmap_read_sb.
Signed-off-by: NeilBrown <neilb@suse.de>
A new attribute directory 'bitmap' in 'md' is created which
contains files for configuring the bitmap.
'location' identifies where the bitmap is, either 'none',
or 'file' or 'sector offset from metadata'.
Writing 'location' can create or remove a bitmap.
Adding a 'file' bitmap this way is not yet supported.
'chunksize' and 'time_base' must be set before 'location'
can be set.
'chunksize' can be set before creating a bitmap, but is
currently always over-ridden by the bitmap superblock.
'time_base' and 'backlog' can be updated at any time.
Signed-off-by: NeilBrown <neilb@suse.de>
Reviewed-by: Andre Noll <maan@systemlinux.org>
safe_delay_store can parse fixed point numbers (for fractions
of a second). We will want to do that for another sysfs
file soon, so factor out the code.
Signed-off-by: NeilBrown <neilb@suse.de>
For md arrays were metadata is managed externally, the kernel does not
know about a superblock so the superblock offset is 0.
If we want to have a write-intent-bitmap near the end of the
devices of such an array, we should support sector_t sized offset.
We need offset be possibly negative for when the bitmap is before
the metadata, so use loff_t instead.
Also add sanity check that bitmap does not overlap with data.
Signed-off-by: NeilBrown <neilb@suse.de>
As bitmap_create and bitmap_destroy already set thread->timeout
as appropriate, there is no need to do it in raid10_quiesce.
There is a possible need to wake the thread after the timeout
has been set low, but it is better to do that where the timeout
is actually set low, in bitmap_create.
Signed-off-by: NeilBrown <neilb@suse.de>
... and into bitmap_info. These are all configuration parameters
that need to be set before the bitmap is created.
Signed-off-by: NeilBrown <neilb@suse.de>
In preparation for making bitmap fields configurable via sysfs,
start tidying up by making a single structure to contain the
configuration fields.
Signed-off-by: NeilBrown <neilb@suse.de>
This will allow us to stop writeout to portions of the array
while they are resynced by someone else - e.g. another node in
a cluster.
Signed-off-by: NeilBrown <neilb@suse.de>
The post-barrier-flush is sent by md as soon as make_request on the
barrier write completes. For raid5, the data might not be in the
per-device queues yet. So for barrier requests, wait for any
pre-reading to be done so that the request will be in the per-device
queues.
We use the 'preread_active' count to check that nothing is still in
the preread phase, and delay the decrement of this count until after
write requests have been submitted to the underlying devices.
Signed-off-by: NeilBrown <neilb@suse.de>
Previously barriers were only supported on RAID1. This is because
other levels requires synchronisation across all devices and so needed
a different approach.
Here is that approach.
When a barrier arrives, we send a zero-length barrier to every active
device. When that completes - and if the original request was not
empty - we submit the barrier request itself (with the barrier flag
cleared) and then submit a fresh load of zero length barriers.
The barrier request itself is asynchronous, but any subsequent
request will block until the barrier completes.
The reason for clearing the barrier flag is that a barrier request is
allowed to fail. If we pass a non-empty barrier through a striping
raid level it is conceivable that part of it could succeed and part
could fail. That would be way too hard to deal with.
So if the first run of zero length barriers succeed, we assume all is
sufficiently well that we send the request and ignore errors in the
second run of barriers.
RAID5 needs extra care as write requests may not have been submitted
to the underlying devices yet. So we flush the stripe cache before
proceeding with the barrier.
Note that the second set of zero-length barriers are submitted
immediately after the original request is submitted. Thus when
a personality finds mddev->barrier to be set during make_request,
it should not return from make_request until the corresponding
per-device request(s) have been queued.
That will be done in later patches.
Signed-off-by: NeilBrown <neilb@suse.de>
Reviewed-by: Andre Noll <maan@systemlinux.org>
If a resync/recovery/check/repair is interrupted for some reason, it
can be useful to know exactly where it got up to.
So in that case, do not clear curr_resync_completed.
Initialise it when starting a resync/recovery/... instead.
Signed-off-by: NeilBrown <neilb@suse.de>
When a 'check' or 'repair' finished we should clear resync_min
so that a future check/repair will cover the whole array (by default).
However if it is interrupted, we should update resync_min to
where we got up to, so that when the check/repair continues it
just does the remainder of the array.
Signed-off-by: NeilBrown <neilb@suse.de>
A write intent bitmap can be removed from an array while the
array is active.
When this happens, all IO is suspended and flushed before the
bitmap is removed.
However it is possible that bitmap_daemon_work is still running to
clear old bits from the bitmap. If it is, it can dereference the
bitmap after it has been freed.
So introduce a new mutex to protect bitmap_daemon_work and get it
before destroying a bitmap.
This is suitable for any current -stable kernel.
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
If the snapshot we are merging became invalid (e.g. it ran out of
space) redirect all I/O directly to the origin device.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Set 'merge_failed' flag if a snapshot fails to merge. Update
snapshot_status() to report "Merge failed" if 'merge_failed' is set.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
s->store->type->prepare_merge returns the number of chunks that can be
copied linearly working backwards from the returned chunk number.
For example, if it returns 3 chunks with old_chunk == 10 and new_chunk
== 20, then chunk 20 can be copied to 10, chunk 19 to 9 and 18 to 8.
Until now kcopyd only copied one chunk at a time. This patch now copies
the full set at once.
Consequently, snapshot_merge_process() needs to delay the merging of all
chunks if any have writes in progress, not just the first chunk in the
region that is to be merged.
snapshot-merge's performance is now comparable to the original
snapshot-origin target.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
When there is one merging snapshot and other non-merging snapshots,
snapshot_merge_process() must make exceptions in the non-merging
snapshots.
Use a sequence count to resolve the race between I/O to chunks that are
about to be merged. The count increases each time an exception
reallocation finishes. Use wait_event() to wait until the count
changes.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Track writes to chunks that are currently being merged and delay merging
a chunk until all writes to that chunk finish.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
While a set of chunks is being merged, any overlapping writes need to be
queued.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Merging is started when origin is resumed and it is stopped when
origin is suspended or when the merging snapshot is destroyed or
errors are detected.
Merging is not yet interlocked with writes: this will be handled in
subsequent patches.
The code relies on callbacks from a private kcopyd thread.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Merging more than one snapshot is not supported, so prevent
this happening.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Sets num_flush_requests=2 to support flushing both the origin and cow
devices used by the snapshot-merge target.
Also, snapshot_ctr() now gets the origin device using FMODE_WRITE if the
target is snapshot-merge (which writes to the origin device).
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
The snapshot-merge target should not allocate new exceptions because the
intent is to merge all of its exceptions as quickly and safely as
possible.
This patch introduces the snapshot-merge mapping function and updates
__origin_write() so that it doesn't allocate exceptions on any snapshots
that are being merged.
If a write request to a merging snapshot device is to be dispatched
directly to the origin (because the chunk is not remapped or was already
merged), snapshot_merge_map() must make exceptions in other snapshots so
calls do_origin().
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
To track the completion of exceptions relating to the same location on
the device, the current code selects one exception as primary_pe, links
the other exceptions to it and uses reference counting to wait until all
the reallocations are complete.
It is considered too complicated to extend this code to handle the new
snapshot-merge target, where sets of non-overlapping chunks would also
need to become linked.
Instead, a simpler (but less efficient) approach is taken. Bios are
linked to one exception. When it completes, bios are simply retried,
and if other related exceptions are still outstanding, they'll get
queued again to wait for another one.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
The snapshot-merge target allows a snapshot to be merged back into the
snapshot's origin device.
One anticipated use of snapshot merging is the rollback of filesystems
to back out problematic system upgrades.
This patch adds snapshot-merge target management to both
dm_snapshot_init() and dm_snapshot_exit(). As an initial place-holder,
snapshot-merge is identical to the snapshot target. Documentation is
provided.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Add functions that decide how many consecutive chunks of snapshot to
merge back into the origin next and to update the metadata afterwards.
prepare_merge provides a pointer to the most recent still-to-be-merged
chunk and returns how many previous ones are consecutive and can be
processed together.
commit_merge removes the nr_merged most-recent chunks permanently from
the exception store. The number must not exceed that returned by
prepare_merge.
Introduce NUM_SNAPSHOT_HDR_CHUNKS to show where the snapshot header
chunk is accounted for.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Move the __chunk_is_tracked() loop into a separate function as we will
also need to call it from the write path in the rare case of conflicting
writes to the same chunk.
Originally introduced in commit a8d41b59f3
("dm snapshot: fix race during exception creation").
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
To support the merging of snapshots back into their origin we need
to trigger exceptions in other snapshots not being merged without
any incoming bio on the origin device. The bio parameter to
__origin_write() becomes optional and the sector needs supplying
separately.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch rejects messages that can generate I/O while the device
itself is suspended.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch adds the exported dm_suspended() function so that targets
can check whether or not they are suspended.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch renames dm_suspended() to dm_suspended_md() and
keeps it internal to dm.
No functional change.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
This patch moves DMF_SUSPENDED flag set before postsuspend.
No one should care about the ordering, because the flag set and
the postsuspend are protected by a single lock, md->suspend_lock,
and all strict flag-checkers take the lock.
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Mike Anderson <andmike@linux.vnet.ibm.com>
Signed-off-by: Alasdair G Kergon <agk@redhat.com>
