Now one page array is allocated for each resync bio, and we can
retrieve page from this table directly.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Now we allocate one page array for managing resync pages, instead
of using bio's vec table to do that, and the old way is very hacky
and won't work any more if multipage bvec is enabled.
The introduced cost is that we need to allocate (128 + 16) * raid_disks
bytes per r1_bio, and it is fine because the inflight r1_bio for
resync shouldn't be much, as pointed by Shaohua.
Also the bio_reset() in raid1_sync_request() is removed because
all bios are freshly new now and not necessary to reset any more.
This patch can be thought as a cleanup too
Suggested-by: Shaohua Li <shli@kernel.org>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
This patch gets each page's reference of each bio for resync,
then r1buf_pool_free() gets simplified a lot.
The same policy has been taken in raid10's buf pool allocation/free
too.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Both raid1 and raid10 share common resync
block size and page count, so move them into md.h.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
All bio_add_page() is for adding one page into resync bio,
which is big enough to hold RESYNC_PAGES pages, and
the current bio_add_page() doesn't check queue limit any more,
so it won't fail at all.
remove unused label (shaohua)
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Change to use bio->__bi_remaining to count number of r1bio attached
to a bio.
See precious raid10 patch for more details.
Like the raid10.c patch, this fixes a bug as nr_queued and nr_pending
used to measure different things, but were being compared.
This patch fixes another bug in that nr_pending previously did not
could write-behind requests, so behind writes could continue while
resync was happening. How that nr_pending counts all r1_bio,
the resync cannot commence until the behind writes have completed.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
When raid1 or raid10 find they will need to allocate a new
r1bio/r10bio, in order to work around a known bad block, they
account for the allocation well before the allocation is
made. This separation makes the correctness less obvious
and requires comments.
The accounting needs to be a little before: before the first
rXbio is submitted, but that is all.
So move the accounting down to where it makes more sense.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Include information about PPL location and size into mdp_superblock_1
and copy it to/from rdev. Because PPL is mutually exclusive with bitmap,
put it in place of 'bitmap_offset'. Add a new flag MD_FEATURE_PPL for
'feature_map', analogically to MD_FEATURE_BITMAP_OFFSET. Add MD_HAS_PPL
to mddev->flags to indicate that PPL is enabled on an array.
Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
raid1.c: fix a trivial typo in comments of freeze_array().
Cc: Jack Wang <jack.wang.usish@gmail.com>
Cc: Guoqing Jiang <gqjiang@suse.com>
Cc: John Stoffel <john@stoffel.org>
Acked-by: Coly Li <colyli@suse.de>
Signed-off-by: Zhilong Liu <zlliu@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Neil Brown pointed out a potential deadlock in raid 10 code with
bio_split/chain. The raid1 code could have the same issue, but recent
barrier rework makes it less likely to happen. The deadlock happens in
below sequence:
1. generic_make_request(bio), this will set current->bio_list
2. raid10_make_request will split bio to bio1 and bio2
3. __make_request(bio1), wait_barrer, add underlayer disk bio to
current->bio_list
4. __make_request(bio2), wait_barrer
If raise_barrier happens between 3 & 4, since wait_barrier runs at 3,
raise_barrier waits for IO completion from 3. And since raise_barrier
sets barrier, 4 waits for raise_barrier. But IO from 3 can't be
dispatched because raid10_make_request() doesn't finished yet.
The solution is to adjust the IO ordering. Quotes from Neil:
"
It is much safer to:
if (need to split) {
split = bio_split(bio, ...)
bio_chain(...)
make_request_fn(split);
generic_make_request(bio);
} else
make_request_fn(mddev, bio);
This way we first process the initial section of the bio (in 'split')
which will queue some requests to the underlying devices. These
requests will be queued in generic_make_request.
Then we queue the remainder of the bio, which will be added to the end
of the generic_make_request queue.
Then we return.
generic_make_request() will pop the lower-level device requests off the
queue and handle them first. Then it will process the remainder
of the original bio once the first section has been fully processed.
"
Note, this only happens in read path. In write path, the bio is flushed to
underlaying disks either by blk flush (from schedule) or offladed to raid1/10d.
It's queued in current->bio_list.
Cc: Coly Li <colyli@suse.de>
Cc: stable@vger.kernel.org (v3.14+, only the raid10 part)
Suggested-by: NeilBrown <neilb@suse.com>
Reviewed-by: Jack Wang <jinpu.wang@profitbricks.com>
Signed-off-by: Shaohua Li <shli@fb.com>
raid1_resize and raid5_resize should also check the
mddev->queue if run underneath dm-raid.
And both set_capacity and revalidate_disk are used in
pers->resize such as raid1, raid10 and raid5. So
move them from personality file to common code.
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
We are going to split <linux/sched/signal.h> out of <linux/sched.h>, which
will have to be picked up from other headers and a couple of .c files.
Create a trivial placeholder <linux/sched/signal.h> file that just
maps to <linux/sched.h> to make this patch obviously correct and
bisectable.
Include the new header in the files that are going to need it.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull md updates from Shaohua Li:
"Mainly fixes bugs and improves performance:
- Improve scalability for raid1 from Coly
- Improve raid5-cache read performance, disk efficiency and IO
pattern from Song and me
- Fix a race condition of disk hotplug for linear from Coly
- A few cleanup patches from Ming and Byungchul
- Fix a memory leak from Neil
- Fix WRITE SAME IO failure from me
- Add doc for raid5-cache from me"
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md: (23 commits)
md/raid1: fix write behind issues introduced by bio_clone_bioset_partial
md/raid1: handle flush request correctly
md/linear: shutup lockdep warnning
md/raid1: fix a use-after-free bug
RAID1: avoid unnecessary spin locks in I/O barrier code
RAID1: a new I/O barrier implementation to remove resync window
md/raid5: Don't reinvent the wheel but use existing llist API
md: fast clone bio in bio_clone_mddev()
md: remove unnecessary check on mddev
md/raid1: use bio_clone_bioset_partial() in case of write behind
md: fail if mddev->bio_set can't be created
block: introduce bio_clone_bioset_partial()
md: disable WRITE SAME if it fails in underlayer disks
md/raid5-cache: exclude reclaiming stripes in reclaim check
md/raid5-cache: stripe reclaim only counts valid stripes
MD: add doc for raid5-cache
Documentation: move MD related doc into a separate dir
md: ensure md devices are freed before module is unloaded.
md/r5cache: improve journal device efficiency
md/r5cache: enable chunk_aligned_read with write back cache
...
There are two issues, introduced by commit 8e58e32(md/raid1: use
bio_clone_bioset_partial() in case of write behind):
- bio_clone_bioset_partial() uses bytes instead of sectors as parameters
- in writebehind mode, we return bio if all !writemostly disk bios finish,
which could happen before writemostly disk bios run. So all
writemostly disk bios should have their bvec. Here we just make sure
all bios are cloned instead of fast cloned.
Reviewed-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
I got a warning triggered in align_to_barrier_unit_end. It's a flush
request so sectors == 0. The flush request happens to work well without
the new barrier patch, but we'd better handle it explictly.
Cc: NeilBrown <neilb@suse.com>
Acked-by: Coly Li <colyli@suse.de>
Signed-off-by: Shaohua Li <shli@fb.com>
Commit fd76863 (RAID1: a new I/O barrier implementation to remove resync
window) introduces a user-after-free bug.
Signed-off-by: Shaohua Li <shli@fb.com>
When I run a parallel reading performan testing on a md raid1 device with
two NVMe SSDs, I observe very bad throughput in supprise: by fio with 64KB
block size, 40 seq read I/O jobs, 128 iodepth, overall throughput is
only 2.7GB/s, this is around 50% of the idea performance number.
The perf reports locking contention happens at allow_barrier() and
wait_barrier() code,
- 41.41% fio [kernel.kallsyms] [k] _raw_spin_lock_irqsave
- _raw_spin_lock_irqsave
+ 89.92% allow_barrier
+ 9.34% __wake_up
- 37.30% fio [kernel.kallsyms] [k] _raw_spin_lock_irq
- _raw_spin_lock_irq
- 100.00% wait_barrier
The reason is, in these I/O barrier related functions,
- raise_barrier()
- lower_barrier()
- wait_barrier()
- allow_barrier()
They always hold conf->resync_lock firstly, even there are only regular
reading I/Os and no resync I/O at all. This is a huge performance penalty.
The solution is a lockless-like algorithm in I/O barrier code, and only
holding conf->resync_lock when it has to.
The original idea is from Hannes Reinecke, and Neil Brown provides
comments to improve it. I continue to work on it, and make the patch into
current form.
In the new simpler raid1 I/O barrier implementation, there are two
wait barrier functions,
- wait_barrier()
Which calls _wait_barrier(), is used for regular write I/O. If there is
resync I/O happening on the same I/O barrier bucket, or the whole
array is frozen, task will wait until no barrier on same barrier bucket,
or the whold array is unfreezed.
- wait_read_barrier()
Since regular read I/O won't interfere with resync I/O (read_balance()
will make sure only uptodate data will be read out), it is unnecessary
to wait for barrier in regular read I/Os, waiting in only necessary
when the whole array is frozen.
The operations on conf->nr_pending[idx], conf->nr_waiting[idx], conf->
barrier[idx] are very carefully designed in raise_barrier(),
lower_barrier(), _wait_barrier() and wait_read_barrier(), in order to
avoid unnecessary spin locks in these functions. Once conf->
nr_pengding[idx] is increased, a resync I/O with same barrier bucket index
has to wait in raise_barrier(). Then in _wait_barrier() if no barrier
raised in same barrier bucket index and array is not frozen, the regular
I/O doesn't need to hold conf->resync_lock, it can just increase
conf->nr_pending[idx], and return to its caller. wait_read_barrier() is
very similar to _wait_barrier(), the only difference is it only waits when
array is frozen. For heavy parallel reading I/Os, the lockless I/O barrier
code almostly gets rid of all spin lock cost.
This patch significantly improves raid1 reading peroformance. From my
testing, a raid1 device built by two NVMe SSD, runs fio with 64KB
blocksize, 40 seq read I/O jobs, 128 iodepth, overall throughput
increases from 2.7GB/s to 4.6GB/s (+70%).
Changelog
V4:
- Change conf->nr_queued[] to atomic_t.
- Define BARRIER_BUCKETS_NR_BITS by (PAGE_SHIFT - ilog2(sizeof(atomic_t)))
V3:
- Add smp_mb__after_atomic() as Shaohua and Neil suggested.
- Change conf->nr_queued[] from atomic_t to int.
- Change conf->array_frozen from atomic_t back to int, and use
READ_ONCE(conf->array_frozen) to check value of conf->array_frozen
in _wait_barrier() and wait_read_barrier().
- In _wait_barrier() and wait_read_barrier(), add a call to
wake_up(&conf->wait_barrier) after atomic_dec(&conf->nr_pending[idx]),
to fix a deadlock between _wait_barrier()/wait_read_barrier and
freeze_array().
V2:
- Remove a spin_lock/unlock pair in raid1d().
- Add more code comments to explain why there is no racy when checking two
atomic_t variables at same time.
V1:
- Original RFC patch for comments.
Signed-off-by: Coly Li <colyli@suse.de>
Cc: Shaohua Li <shli@fb.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Guoqing Jiang <gqjiang@suse.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Signed-off-by: Shaohua Li <shli@fb.com>
'Commit 79ef3a8aa1 ("raid1: Rewrite the implementation of iobarrier.")'
introduces a sliding resync window for raid1 I/O barrier, this idea limits
I/O barriers to happen only inside a slidingresync window, for regular
I/Os out of this resync window they don't need to wait for barrier any
more. On large raid1 device, it helps a lot to improve parallel writing
I/O throughput when there are background resync I/Os performing at
same time.
The idea of sliding resync widow is awesome, but code complexity is a
challenge. Sliding resync window requires several variables to work
collectively, this is complexed and very hard to make it work correctly.
Just grep "Fixes: 79ef3a8aa1" in kernel git log, there are 8 more patches
to fix the original resync window patch. This is not the end, any further
related modification may easily introduce more regreassion.
Therefore I decide to implement a much simpler raid1 I/O barrier, by
removing resync window code, I believe life will be much easier.
The brief idea of the simpler barrier is,
- Do not maintain a global unique resync window
- Use multiple hash buckets to reduce I/O barrier conflicts, regular
I/O only has to wait for a resync I/O when both them have same barrier
bucket index, vice versa.
- I/O barrier can be reduced to an acceptable number if there are enough
barrier buckets
Here I explain how the barrier buckets are designed,
- BARRIER_UNIT_SECTOR_SIZE
The whole LBA address space of a raid1 device is divided into multiple
barrier units, by the size of BARRIER_UNIT_SECTOR_SIZE.
Bio requests won't go across border of barrier unit size, that means
maximum bio size is BARRIER_UNIT_SECTOR_SIZE<<9 (64MB) in bytes.
For random I/O 64MB is large enough for both read and write requests,
for sequential I/O considering underlying block layer may merge them
into larger requests, 64MB is still good enough.
Neil also points out that for resync operation, "we want the resync to
move from region to region fairly quickly so that the slowness caused
by having to synchronize with the resync is averaged out over a fairly
small time frame". For full speed resync, 64MB should take less then 1
second. When resync is competing with other I/O, it could take up a few
minutes. Therefore 64MB size is fairly good range for resync.
- BARRIER_BUCKETS_NR
There are BARRIER_BUCKETS_NR buckets in total, which is defined by,
#define BARRIER_BUCKETS_NR_BITS (PAGE_SHIFT - 2)
#define BARRIER_BUCKETS_NR (1<<BARRIER_BUCKETS_NR_BITS)
this patch makes the bellowed members of struct r1conf from integer
to array of integers,
- int nr_pending;
- int nr_waiting;
- int nr_queued;
- int barrier;
+ int *nr_pending;
+ int *nr_waiting;
+ int *nr_queued;
+ int *barrier;
number of the array elements is defined as BARRIER_BUCKETS_NR. For 4KB
kernel space page size, (PAGE_SHIFT - 2) indecates there are 1024 I/O
barrier buckets, and each array of integers occupies single memory page.
1024 means for a request which is smaller than the I/O barrier unit size
has ~0.1% chance to wait for resync to pause, which is quite a small
enough fraction. Also requesting single memory page is more friendly to
kernel page allocator than larger memory size.
- I/O barrier bucket is indexed by bio start sector
If multiple I/O requests hit different I/O barrier units, they only need
to compete I/O barrier with other I/Os which hit the same I/O barrier
bucket index with each other. The index of a barrier bucket which a
bio should look for is calculated by sector_to_idx() which is defined
in raid1.h as an inline function,
static inline int sector_to_idx(sector_t sector)
{
return hash_long(sector >> BARRIER_UNIT_SECTOR_BITS,
BARRIER_BUCKETS_NR_BITS);
}
Here sector_nr is the start sector number of a bio.
- Single bio won't go across boundary of a I/O barrier unit
If a request goes across boundary of barrier unit, it will be split. A
bio may be split in raid1_make_request() or raid1_sync_request(), if
sectors returned by align_to_barrier_unit_end() is smaller than
original bio size.
Comparing to single sliding resync window,
- Currently resync I/O grows linearly, therefore regular and resync I/O
will conflict within a single barrier units. So the I/O behavior is
similar to single sliding resync window.
- But a barrier unit bucket is shared by all barrier units with identical
barrier uinit index, the probability of conflict might be higher
than single sliding resync window, in condition that writing I/Os
always hit barrier units which have identical barrier bucket indexs with
the resync I/Os. This is a very rare condition in real I/O work loads,
I cannot imagine how it could happen in practice.
- Therefore we can achieve a good enough low conflict rate with much
simpler barrier algorithm and implementation.
There are two changes should be noticed,
- In raid1d(), I change the code to decrease conf->nr_pending[idx] into
single loop, it looks like this,
spin_lock_irqsave(&conf->device_lock, flags);
conf->nr_queued[idx]--;
spin_unlock_irqrestore(&conf->device_lock, flags);
This change generates more spin lock operations, but in next patch of
this patch set, it will be replaced by a single line code,
atomic_dec(&conf->nr_queueud[idx]);
So we don't need to worry about spin lock cost here.
- Mainline raid1 code split original raid1_make_request() into
raid1_read_request() and raid1_write_request(). If the original bio
goes across an I/O barrier unit size, this bio will be split before
calling raid1_read_request() or raid1_write_request(), this change
the code logic more simple and clear.
- In this patch wait_barrier() is moved from raid1_make_request() to
raid1_write_request(). In raid_read_request(), original wait_barrier()
is replaced by raid1_read_request().
The differnece is wait_read_barrier() only waits if array is frozen,
using different barrier function in different code path makes the code
more clean and easy to read.
Changelog
V4:
- Add alloc_r1bio() to remove redundant r1bio memory allocation code.
- Fix many typos in patch comments.
- Use (PAGE_SHIFT - ilog2(sizeof(int))) to define BARRIER_BUCKETS_NR_BITS.
V3:
- Rebase the patch against latest upstream kernel code.
- Many fixes by review comments from Neil,
- Back to use pointers to replace arraries in struct r1conf
- Remove total_barriers from struct r1conf
- Add more patch comments to explain how/why the values of
BARRIER_UNIT_SECTOR_SIZE and BARRIER_BUCKETS_NR are decided.
- Use get_unqueued_pending() to replace get_all_pendings() and
get_all_queued()
- Increase bucket number from 512 to 1024
- Change code comments format by review from Shaohua.
V2:
- Use bio_split() to split the orignal bio if it goes across barrier unit
bounday, to make the code more simple, by suggestion from Shaohua and
Neil.
- Use hash_long() to replace original linear hash, to avoid a possible
confilict between resync I/O and sequential write I/O, by suggestion from
Shaohua.
- Add conf->total_barriers to record barrier depth, which is used to
control number of parallel sync I/O barriers, by suggestion from Shaohua.
- In V1 patch the bellowed barrier buckets related members in r1conf are
allocated in memory page. To make the code more simple, V2 patch moves
the memory space into struct r1conf, like this,
- int nr_pending;
- int nr_waiting;
- int nr_queued;
- int barrier;
+ int nr_pending[BARRIER_BUCKETS_NR];
+ int nr_waiting[BARRIER_BUCKETS_NR];
+ int nr_queued[BARRIER_BUCKETS_NR];
+ int barrier[BARRIER_BUCKETS_NR];
This change is by the suggestion from Shaohua.
- Remove some inrelavent code comments, by suggestion from Guoqing.
- Add a missing wait_barrier() before jumping to retry_write, in
raid1_make_write_request().
V1:
- Original RFC patch for comments
Signed-off-by: Coly Li <colyli@suse.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Guoqing Jiang <gqjiang@suse.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Signed-off-by: Shaohua Li <shli@fb.com>
Firstly bio_clone_mddev() is used in raid normal I/O and isn't
in resync I/O path.
Secondly all the direct access to bvec table in raid happens on
resync I/O except for write behind of raid1, in which we still
use bio_clone() for allocating new bvec table.
So this patch replaces bio_clone() with bio_clone_fast()
in bio_clone_mddev().
Also kill bio_clone_mddev() and call bio_clone_fast() directly, as
suggested by Christoph Hellwig.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Write behind need to replace pages in bio's bvecs, and we have
to clone a fresh bio with new bvec table, so use the introduced
bio_clone_bioset_partial() for it.
For other bio_clone_mddev() cases, we will use fast clone since
they don't need to touch bvec table.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Shaohua Li <shli@fb.com>
We will want to have struct backing_dev_info allocated separately from
struct request_queue. As the first step add pointer to backing_dev_info
to request_queue and convert all users touching it. No functional
changes in this patch.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Jens Axboe <axboe@fb.com>
No need for the local variables, the bio is still live and we can just
assign the bits we want directly. Make me wonder why we can't assign
all the bio flags to start with.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Commit 6995f0b (md: takeover should clear unrelated bits) clear
unrelated bits, but it's quite fragile. To avoid error in the future,
define a macro for unsupported mddev flags for each raid type and use it
to clear unsupported mddev flags. This should be less error-prone.
Suggested-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Refactor raid1_make_request to make read and write code in their own
functions to clean up the code.
Signed-off-by: Robert LeBlanc <robert@leblancnet.us>
Signed-off-by: Shaohua Li <shli@fb.com>
The mddev->flags are used for different purposes. There are a lot of
places we check/change the flags without masking unrelated flags, we
could check/change unrelated flags. These usage are most for superblock
write, so spearate superblock related flags. This should make the code
clearer and also fix real bugs.
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
When we change level from raid1 to raid5, the MD_FAILFAST_SUPPORTED bit
will be accidentally set, but raid5 doesn't support it. The same is true
for the MD_HAS_JOURNAL bit.
Fix: 46533ff (md: Use REQ_FAILFAST_* on metadata writes where appropriate)
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
When writing to a fastfail device we use MD_FASTFAIL unless
it is the only device being written to.
For resync/recovery, assume there was a working device to
read from so always use REQ_FASTFAIL_DEV.
If a write for resync/recovery fails, we just fail the
device - there is not much else to do.
If a normal failfast write fails, but the device cannot be
failed (must be only one left), we queue for write error
handling. This will call narrow_write_error() to retry the
write synchronously and without any FAILFAST flags.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
If a device is marked FailFast and it is not the only device
we can read from, we mark the bio with REQ_FAILFAST_* flags.
If this does fail, we don't try read repair but just allow
failure. If it was the last device it doesn't fail of
course, so the retry happens on the same device - this time
without FAILFAST. A subsequent failure will not retry but
will just pass up the error.
During resync we may use FAILFAST requests and on a failure
we will simply use the other device(s).
During recovery we will only use FAILFAST in the unusual
case were there are multiple places to read from - i.e. if
there are > 2 devices. If we get a failure we will fail the
device and complete the resync/recovery with remaining
devices.
The new R1BIO_FailFast flag is set on read reqest to suggest
the a FAILFAST request might be acceptable. The rdev needs
to have FailFast set as well for the read to actually use
REQ_FAILFAST_*.
We need to know there are at least two working devices
before we can set R1BIO_FailFast, so we mustn't stop looking
at the first device we find. So the "min_pending == 0"
handling to not exit early, but too always choose the
best_pending_disk if min_pending == 0.
The spinlocked region in raid1_error() in enlarged to ensure
that if two bios, reading from two different devices, fail
at the same time, then there is no risk that both devices
will be marked faulty, leaving zero "In_sync" devices.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
This can only be supported on personalities which ensure
that md_error() never causes an array to enter the 'failed'
state. i.e. if marking a device Faulty would cause some
data to be inaccessible, the device is status is left as
non-Faulty. This is true for RAID1 and RAID10.
If we get a failure writing metadata but the device doesn't
fail, it must be the last device so we re-write without
FAILFAST to improve chance of success. We also flag the
device as LastDev so that future metadata updates don't
waste time on failfast writes.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Both raid1 and raid10 will sometimes delay handling an IO request,
such as when resync is happening or there are too many requests queued.
Add some blktrace messsages so we can see when that is happening when
looking for performance artefacts.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
The block tracing infrastructure (accessed with blktrace/blkparse)
supports the tracing of mapping bios from one device to another.
This is currently used when a bio in a partition is mapped to the
whole device, when bios are mapped by dm, and for mapping in md/raid5.
Other md personalities do not include this tracing yet, so add it.
When a read-error is detected we redirect the request to a different device.
This could justifiably be seen as a new mapping for the originial bio,
or a secondary mapping for the bio that errors. This patch uses
the second option.
When md is used under dm-raid, the mappings are not traced as we do
not have access to the block device number of the parent.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
While performing a resync/recovery, raid1 divides the
array space into three regions:
- before the resync
- at or shortly after the resync point
- much further ahead of the resync point.
Write requests to the first or third do not need to wait. Write
requests to the middle region do need to wait if resync requests are
pending.
If there are any active write requests in the middle region, resync
will wait for them.
Due to an accounting error, there is a small range of addresses,
between conf->next_resync and conf->start_next_window, where write
requests will *not* be blocked, but *will* be counted in the middle
region. This can effectively block resync indefinitely if filesystem
writes happen repeatedly to this region.
As ->next_window_requests is incremented when the sector is after
conf->start_next_window + NEXT_NORMALIO_DISTANCE
the same boundary should be used for determining when write requests
should wait.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
When writing to an array with a bitmap enabled, the writes are grouped
in batches which are preceded by an update to the bitmap.
It is quite likely if that a drive develops a problem which is not
media related, that the bitmap write will be the first to report an
error and cause the device to be marked faulty (as the bitmap write is
at the start of a batch).
In this case, there is point submiting the subsequent writes to the
failed device - that just wastes times.
So re-check the Faulty state of a device before submitting a
delayed write.
This requires that we keep the 'rdev', rather than the 'bdev' in the
bio, then swap in the bdev just before final submission.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
If write is the first operation on a disk and it happens not to be
aligned to page size, block layer sends read request first. If read
operation fails, the disk is set as failed as no attempt to fix the
error is made because array is in auto-readonly mode. Similarily, the
disk is set as failed for read-only array.
Take the same approach as in raid10. Don't fail the disk if array is in
readonly or auto-readonly mode. Try to redirect the request first and if
unsuccessful, return a read error.
Signed-off-by: Tomasz Majchrzak <tomasz.majchrzak@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
If a write error occurs, raid1 will try to rewrite the bio in small
chunk size. If the rewrite fails, raid1 will record the error in bad
block. narrow_write_error will always use WRITE for the bio, but
actually it could be a discard. Since discard bio hasn't payload, write
the bio will cause different issues. But discard error isn't fatal, we
can safely ignore it. This is what this patch does.
This issue should exist since discard is added, but only exposed with
recent arbitrary bio size feature.
Reported-and-tested-by: Sitsofe Wheeler <sitsofe@gmail.com>
Cc: stable@vger.kernel.org (v3.6)
Signed-off-by: Shaohua Li <shli@fb.com>
bio_free_pages is introduced in commit 1dfa0f68c0
("block: add a helper to free bio bounce buffer pages"),
we can reuse the func in other modules after it was
imported.
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Jens Axboe <axboe@fb.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Shaohua Li <shli@fb.com>
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Acked-by: Kent Overstreet <kent.overstreet@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Since commit 63a4cc2486, bio->bi_rw contains flags in the lower
portion and the op code in the higher portions. This means that
old code that relies on manually setting bi_rw is most likely
going to be broken. Instead of letting that brokeness linger,
rename the member, to force old and out-of-tree code to break
at compile time instead of at runtime.
No intended functional changes in this commit.
Signed-off-by: Jens Axboe <axboe@fb.com>
These two are confusing leftover of the old world order, combining
values of the REQ_OP_ and REQ_ namespaces. For callers that don't
special case we mostly just replace bi_rw with bio_data_dir or
op_is_write, except for the few cases where a switch over the REQ_OP_
values makes more sense. Any check for READA is replaced with an
explicit check for REQ_RAHEAD. Also remove the READA alias for
REQ_RAHEAD.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Mike Christie <mchristi@redhat.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Every time a device is removed with ->hot_remove_disk() a synchronize_rcu() call is made
which can delay several milliseconds in some case.
If lots of devices fail at once - as could happen with a large RAID10 where one set
of devices are removed all at once - these delays can add up to be very inconcenient.
As failure is not reversible we can check for that first, setting a
separate flag if it is found, and then all synchronize_rcu() once for
all the flagged devices. Then ->hot_remove_disk() function can skip the
synchronize_rcu() step if the flag is set.
fix build error(Shaohua)
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Since remove_and_add_spares() was added to hot_remove_disk() it has
been possible for an rdev to be hot-removed while fix_read_error()
was running, so we need to be more careful, and take a reference to
the rdev while performing IO.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
'mirror' is only used to find 'rdev', several times.
So just find 'rdev' once, and use it instead.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Both functions use conf->mirrors[mirror].rdev several times, so
improve readability by storing this in a local variable.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Re-checking the faulty flag here brings no value.
The comment about "risk" refers to the risk that the device could
be in the process of being removed by ->hot_remove_disk().
However providing that the ->nr_pending count is incremented inside
an rcu_read_locked() region, there is no risk of that happening.
This is because the rdev pointer (in the personalities array) is set
to NULL before synchronize_rcu(), and ->nr_pending is tested
afterwards. If the rcu_read_locked region happens before the
synchronize_rcu(), the test will see that nr_pending has been incremented.
If it happens afterwards, the rdev pointer will be NULL so there is nothing
to increment.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
A performance drop of mkfs has been observed on RAID10 during resync
since commit 09314799e4 ("md: remove 'go_faster' option from
->sync_request()"). Resync sends so many IOs it slows down non-resync
IOs significantly (few times). Add a short delay to a resync. The
previous long sleep (1s) has proven unnecessary, even very short delay
brings performance right.
The change also applied to raid1. The problem has not been observed on
raid1, however it shares barriers code with raid10 so it might be an
issue for some setup too.
Suggested-by: NeilBrown <neilb@suse.com>
Link: http://lkml.kernel.org/r/20160609134555.GA9104@proton.igk.intel.com
Signed-off-by: Tomasz Majchrzak <tomasz.majchrzak@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Instead of overloading the discard support with the REQ_SECURE flag.
Use the opportunity to rename the queue flag as well, and remove the
dead checks for this flag in the RAID 1 and RAID 10 drivers that don't
claim support for secure erase.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
To avoid confusion between REQ_OP_FLUSH, which is handled by
request_fn drivers, and upper layers requesting the block layer
perform a flush sequence along with possibly a WRITE, this patch
renames REQ_FLUSH to REQ_PREFLUSH.
Signed-off-by: Mike Christie <mchristi@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Separate the op from the rq_flag_bits and have md
set/get the bio using bio_set_op_attrs/bio_op.
Signed-off-by: Mike Christie <mchristi@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This has callers of submit_bio/submit_bio_wait set the bio->bi_rw
instead of passing it in. This makes that use the same as
generic_make_request and how we set the other bio fields.
Signed-off-by: Mike Christie <mchristi@redhat.com>
Fixed up fs/ext4/crypto.c
Signed-off-by: Jens Axboe <axboe@fb.com>
Some code waits for a metadata update by:
1. flagging that it is needed (MD_CHANGE_DEVS or MD_CHANGE_CLEAN)
2. setting MD_CHANGE_PENDING and waking the management thread
3. waiting for MD_CHANGE_PENDING to be cleared
If the first two are done without locking, the code in md_update_sb()
which checks if it needs to repeat might test if an update is needed
before step 1, then clear MD_CHANGE_PENDING after step 2, resulting
in the wait returning early.
So make sure all places that set MD_CHANGE_PENDING are atomicial, and
bit_clear_unless (suggested by Neil) is introduced for the purpose.
Cc: Martin Kepplinger <martink@posteo.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: <linux-kernel@vger.kernel.org>
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
If first_bad == this_sector when we get the WriteMostly disk
in read_balance(), valid disk will be returned with zero
max_sectors. It'll lead to a dead loop in make_request(), and
OOM will happen because of endless allocation of struct bio.
Since we can't get data from this disk in this case, so
continue for another disk.
Signed-off-by: Wei Fang <fangwei1@huawei.com>
Signed-off-by: Shaohua Li <shli@fb.com>
If raid1d is handling a mix of read and write errors, handle_read_error's
call to freeze_array can get stuck.
This can happen because, though the bio_end_io_list is initially drained,
writes can be added to it via handle_write_finished as the retry_list
is processed. These writes contribute to nr_pending but are not included
in nr_queued.
If a later entry on the retry_list triggers a call to handle_read_error,
freeze array hangs waiting for nr_pending == nr_queued+extra. The writes
on the bio_end_io_list aren't included in nr_queued so the condition will
never be satisfied.
To prevent the hang, include bio_end_io_list writes in nr_queued.
There's probably a better way to handle decrementing nr_queued, but this
seemed like the safest way to avoid breaking surrounding code.
I'm happy to supply the script I used to repro this hang.
Fixes: 55ce74d4bfe1b(md/raid1: ensure device failure recorded before write request returns.)
Cc: stable@vger.kernel.org (v4.3+)
Signed-off-by: Nate Dailey <nate.dailey@stratus.com>
Signed-off-by: Shaohua Li <shli@fb.com>
Since bitmap_start_sync will not return until
sync_blocks is not less than PAGE_SIZE>>9, so
the BUG_ON is not needed anymore.
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
It is not safe for an integrity profile to be changed while i/o is
in-flight in the queue. Prevent adding new disks or otherwise online
spares to an array if the device has an incompatible integrity profile.
The original change to the blk_integrity_unregister implementation in
md, commmit c7bfced9a6 "md: suspend i/o during runtime
blk_integrity_unregister" introduced an immediate hang regression.
This policy of disallowing changes the integrity profile once one has
been established is shared with DM.
Here is an abbreviated log from a test run that:
1/ Creates a degraded raid1 with an integrity-enabled device (pmem0s) [ 59.076127]
2/ Tries to add an integrity-disabled device (pmem1m) [ 90.489209]
3/ Retries with an integrity-enabled device (pmem1s) [ 205.671277]
[ 59.076127] md/raid1:md0: active with 1 out of 2 mirrors
[ 59.078302] md: data integrity enabled on md0
[..]
[ 90.489209] md0: incompatible integrity profile for pmem1m
[..]
[ 205.671277] md: super_written gets error=-5
[ 205.677386] md/raid1:md0: Disk failure on pmem1m, disabling device.
[ 205.677386] md/raid1:md0: Operation continuing on 1 devices.
[ 205.683037] RAID1 conf printout:
[ 205.684699] --- wd:1 rd:2
[ 205.685972] disk 0, wo:0, o:1, dev:pmem0s
[ 205.687562] disk 1, wo:1, o:1, dev:pmem1s
[ 205.691717] md: recovery of RAID array md0
Fixes: c7bfced9a6 ("md: suspend i/o during runtime blk_integrity_unregister")
Cc: <stable@vger.kernel.org>
Cc: Mike Snitzer <snitzer@redhat.com>
Reported-by: NeilBrown <neilb@suse.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Two major components to this update.
1/ the clustered-raid1 support from SUSE is nearly
complete. There are a few outstanding issues being
worked on. Maybe half a dozen patches will bring
this to a usable state.
2/ The first stage of journalled-raid5 support from
Facebook makes an appearance. With a journal
device configured (typically NVRAM or SSD), the
"RAID5 write hole" should be closed - a crash
during degraded operations cannot result in data
corruption.
The next stage will be to use the journal as a
write-behind cache so that latency can be reduced
and in some cases throughput increased by
performing more full-stripe writes.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2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=TD5o
-----END PGP SIGNATURE-----
Merge tag 'md/4.4' of git://neil.brown.name/md
Pull md updates from Neil Brown:
"Two major components to this update.
1) The clustered-raid1 support from SUSE is nearly complete. There
are a few outstanding issues being worked on. Maybe half a dozen
patches will bring this to a usable state.
2) The first stage of journalled-raid5 support from Facebook makes an
appearance. With a journal device configured (typically NVRAM or
SSD), the "RAID5 write hole" should be closed - a crash during
degraded operations cannot result in data corruption.
The next stage will be to use the journal as a write-behind cache
so that latency can be reduced and in some cases throughput
increased by performing more full-stripe writes.
* tag 'md/4.4' of git://neil.brown.name/md: (66 commits)
MD: when RAID journal is missing/faulty, block RESTART_ARRAY_RW
MD: set journal disk ->raid_disk
MD: kick out journal disk if it's not fresh
raid5-cache: start raid5 readonly if journal is missing
MD: add new bit to indicate raid array with journal
raid5-cache: IO error handling
raid5: journal disk can't be removed
raid5-cache: add trim support for log
MD: fix info output for journal disk
raid5-cache: use bio chaining
raid5-cache: small log->seq cleanup
raid5-cache: new helper: r5_reserve_log_entry
raid5-cache: inline r5l_alloc_io_unit into r5l_new_meta
raid5-cache: take rdev->data_offset into account early on
raid5-cache: refactor bio allocation
raid5-cache: clean up r5l_get_meta
raid5-cache: simplify state machine when caches flushes are not needed
raid5-cache: factor out a helper to run all stripes for an I/O unit
raid5-cache: rename flushed_ios to finished_ios
raid5-cache: free I/O units earlier
...
Pull block integrity updates from Jens Axboe:
""This is the joint work of Dan and Martin, cleaning up and improving
the support for block data integrity"
* 'for-4.4/integrity' of git://git.kernel.dk/linux-block:
block, libnvdimm, nvme: provide a built-in blk_integrity nop profile
block: blk_flush_integrity() for bio-based drivers
block: move blk_integrity to request_queue
block: generic request_queue reference counting
nvme: suspend i/o during runtime blk_integrity_unregister
md: suspend i/o during runtime blk_integrity_unregister
md, dm, scsi, nvme, libnvdimm: drop blk_integrity_unregister() at shutdown
block: Inline blk_integrity in struct gendisk
block: Export integrity data interval size in sysfs
block: Reduce the size of struct blk_integrity
block: Consolidate static integrity profile properties
block: Move integrity kobject to struct gendisk
To incorporate --grow feature executed on one node, other nodes need to
acknowledge the change in number of disks. Call update_raid_disks()
to update internal data structures.
This leads to call check_reshape() -> md_allow_write() -> md_update_sb(),
this results in a deadlock. This is done so it can safely allocate memory
(which might trigger writeback which might write to raid1). This is
not required for md with a bitmap.
In the clustered case, we don't perform md_update_sb() in md_allow_write(),
but in do_md_run(). Also we disable safemode for clustered mode.
mddev->recovery_cp need not be set in check_sb_changes() because this
is required only when a node reads another node's bitmap. mddev->recovery_cp
(which is read from sb->resync_offset), is set only if mddev is in_sync.
Since we disabled safemode, in_sync is set to zero.
In a clustered environment, the MD may not be in sync because another
node could be writing to it. So make sure that in_sync is not set in
case of clustered node in __md_stop_writes().
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.com>
When a write fails and a bad-block-list is present, we can
update the bad-block-list instead of writing the data. If
this succeeds then it is OK clear the relevant bitmap-bit as
no further 'sync' of the block is needed.
However if writing the bad-block-list fails then we need to
treat the write as failed and particularly must not clear
the bitmap bit. Otherwise the device can be re-added (after
any hardware connection issues are resolved) and because the
relevant bit in the bitmap is clear, that block will not be
resynced. This leads to data corruption.
We already delay the final bio_endio() on the write until
the bad-block-list is written so that when the write
returns: either that data is safe, the bad-block record is
safe, or the fact that the device is faulty is safe.
However we *don't* delay the clearing of the bitmap, so the
bitmap bit can be recorded as cleared before we know if the
bad-block-list was written safely.
So: delay that until the write really is safe.
i.e. move the call to close_write() until just before
calling bio_endio(), and recheck the 'is array degraded'
status before making that call.
This bug goes back to v3.1 when bad-block-lists were
introduced, though it only affects arrays created with
mdadm-3.3 or later as only those have bad-block lists.
Backports will require at least
Commit: 55ce74d4bf ("md/raid1: ensure device failure recorded before write request returns.")
as well. I'll send that to 'stable' separately.
Note that of the two tests of R1BIO_WriteError that this
patch adds, the first is certain to fail and the second is
certain to succeed. However doing it this way makes the
patch more obviously correct. I will tidy the code up in a
future merge window.
Reported-and-tested-by: Nate Dailey <nate.dailey@stratus.com>
Cc: Jes Sorensen <Jes.Sorensen@redhat.com>
Fixes: cd5ff9a16f ("md/raid1: Handle write errors by updating badblock log.")
Signed-off-by: NeilBrown <neilb@suse.com>
Synchronize pending i/o against a change in the integrity profile to
avoid the possibility of spurious integrity errors. Given linear_add()
is suspending the mddev before manipulating the mddev, do the same for
the other personalities.
Acked-by: NeilBrown <neilb@suse.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This was introduced with 9e882242c6
which changed the return value of submit_bio_wait() to return != 0 on
error, but didn't update the caller accordingly.
Fixes: 9e882242c6 ("block: Add submit_bio_wait(), remove from md")
Cc: stable@vger.kernel.org (v3.10)
Reported-by: Bill Kuzeja <William.Kuzeja@stratus.com>
Signed-off-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
md-cluster: A better way for METADATA_UPDATED processing
The processing of METADATA_UPDATED message is too simple and prone to
errors. Besides, it would not update the internal data structures as
required.
This set of patches reads the superblock from one of the device of the MD
and checks for changes in the in-memory data structures. If there is a change,
it performs the necessary actions to keep the internal data structures
as it would be in the primary node.
An example is if a devices turns faulty. The algorithm is:
1. The initiator node marks the device as faulty and updates the superblock
2. The initiator node sends METADATA_UPDATED with an advisory device number to the rest of the nodes.
3. The receiving node on receiving the METADATA_UPDATED message
3.1 Reads the superblock
3.2 Detects a device has failed by comparing with memory structure
3.3 Calls the necessary functions to record the failure and get the device out of the active array.
3.4 Acknowledges the message.
The patch series also fixes adding the disk which was impacted because of
the changes.
Patches can also be found at
https://github.com/goldwynr/linux branch md-next
Changes since V2:
- Fix status synchrnoization after --add and --re-add operations
- Included Guoqing's patches on endian correctness, zeroing cmsg etc
- Restructure add_new_disk() and cancel()
Resync or recovery must be performed by only one node at a time.
A DLM lock resource, resync_lockres provides the mutual exclusion
so that only one node performs the recovery/resync at a time.
If a node is unable to get the resync_lockres, because recovery is
being performed by another node, it set MD_RECOVER_NEEDED so as
to schedule recovery in the future.
Remove the debug message in resync_info_update()
used during development.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
md_reload_sb is too simplistic and it explicitly needs to determine
the changes made by the writing node. However, there are multiple areas
where a simple reload could fail.
Instead, read the superblock of one of the "good" rdevs and update
the necessary information:
- read the superblock into a newly allocated page, by temporarily
swapping out rdev->sb_page and calling ->load_super.
- if that fails return
- if it succeeds, call check_sb_changes
1. iterates over list of active devices and checks the matching
dev_roles[] value.
If that is 'faulty', the device must be marked as faulty
- call md_error to mark the device as faulty. Make sure
not to set CHANGE_DEVS and wakeup mddev->thread or else
it would initiate a resync process, which is the responsibility
of the "primary" node.
- clear the Blocked bit
- Call remove_and_add_spares() to hot remove the device.
If the device is 'spare':
- call remove_and_add_spares() to get the number of spares
added in this operation.
- Reduce mddev->degraded to mark the array as not degraded.
2. reset recovery_cp
- read the rest of the rdevs to update recovery_offset. If recovery_offset
is equal to MaxSector, call spare_active() to set it In_sync
This required that recovery_offset be initialized to MaxSector, as
opposed to zero so as to communicate the end of sync for a rdev.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Suspending the entire device for resync could take too long. Resync
in small chunks.
cluster's resync window (32M) is maintained in r1conf as
cluster_sync_low and cluster_sync_high and processed in
raid1's sync_request(). If the current resync is outside the cluster
resync window:
1. Set the cluster_sync_low to curr_resync_completed.
2. Check if the sync will fit in the new window, if not issue a
wait_barrier() and set cluster_sync_low to sector_nr.
3. Set cluster_sync_high to cluster_sync_low + resync_window.
4. Send a message to all nodes so they may add it in their suspension
list.
bitmap_cond_end_sync is modified to allow to force a sync inorder
to get the curr_resync_completed uptodate with the sector passed.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Signed-off-by: NeilBrown <neilb@suse.de>
The commit 55ce74d4bf (md/raid1: ensure
device failure recorded before write request returns) is causing crash in
the LVM2 testsuite test shell/lvchange-raid.sh. For me the crash is 100%
reproducible.
The reason for the crash is that the newly added code in raid1d moves the
list from conf->bio_end_io_list to tmp, then tests if tmp is non-empty and
then incorrectly pops the bio from conf->bio_end_io_list (which is empty
because the list was alrady moved).
Raid-10 has a similar bug.
Kernel Fault: Code=15 regs=000000006ccb8640 (Addr=0000000100000000)
CPU: 3 PID: 1930 Comm: mdX_raid1 Not tainted 4.2.0-rc5-bisect+ #35
task: 000000006cc1f258 ti: 000000006ccb8000 task.ti: 000000006ccb8000
YZrvWESTHLNXBCVMcbcbcbcbOGFRQPDI
PSW: 00001000000001001111111000001111 Not tainted
r00-03 000000ff0804fe0f 000000001059d000 000000001059f818 000000007f16be38
r04-07 000000001059d000 000000007f16be08 0000000000200200 0000000000000001
r08-11 000000006ccb8260 000000007b7934d0 0000000000000001 0000000000000000
r12-15 000000004056f320 0000000000000000 0000000000013dd0 0000000000000000
r16-19 00000000f0d00ae0 0000000000000000 0000000000000000 0000000000000001
r20-23 000000000800000f 0000000042200390 0000000000000000 0000000000000000
r24-27 0000000000000001 000000000800000f 000000007f16be08 000000001059d000
r28-31 0000000100000000 000000006ccb8560 000000006ccb8640 0000000000000000
sr00-03 0000000000249800 0000000000000000 0000000000000000 0000000000249800
sr04-07 0000000000000000 0000000000000000 0000000000000000 0000000000000000
IASQ: 0000000000000000 0000000000000000 IAOQ: 000000001059f61c 000000001059f620
IIR: 0f8010c6 ISR: 0000000000000000 IOR: 0000000100000000
CPU: 3 CR30: 000000006ccb8000 CR31: 0000000000000000
ORIG_R28: 000000001059d000
IAOQ[0]: call_bio_endio+0x34/0x1a8 [raid1]
IAOQ[1]: call_bio_endio+0x38/0x1a8 [raid1]
RP(r2): raid_end_bio_io+0x88/0x168 [raid1]
Backtrace:
[<000000001059f818>] raid_end_bio_io+0x88/0x168 [raid1]
[<00000000105a4f64>] raid1d+0x144/0x1640 [raid1]
[<000000004017fd5c>] kthread+0x144/0x160
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Fixes: 55ce74d4bf ("md/raid1: ensure device failure recorded before write request returns.")
Fixes: 95af587e95 ("md/raid10: ensure device failure recorded before write request returns.")
Signed-off-by: NeilBrown <neilb@suse.com>
close_sync() needs to set conf->next_resync to a large, but safe value
below MaxSector and use it to determine whether or not to set
start_next_window in wait_barrier()
Solution suggested by Neil Brown.
Reported-by: Nate Dailey <nate.dailey@stratus.com>
Tested-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Remove unneeded NULL test.
The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)
// <smpl>
@@ expression x; @@
-if (x != NULL)
\(kmem_cache_destroy\|mempool_destroy\|dma_pool_destroy\)(x);
// </smpl>
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: NeilBrown <neilb@suse.com>
Pull core block updates from Jens Axboe:
"This first core part of the block IO changes contains:
- Cleanup of the bio IO error signaling from Christoph. We used to
rely on the uptodate bit and passing around of an error, now we
store the error in the bio itself.
- Improvement of the above from myself, by shrinking the bio size
down again to fit in two cachelines on x86-64.
- Revert of the max_hw_sectors cap removal from a revision again,
from Jeff Moyer. This caused performance regressions in various
tests. Reinstate the limit, bump it to a more reasonable size
instead.
- Make /sys/block/<dev>/queue/discard_max_bytes writeable, by me.
Most devices have huge trim limits, which can cause nasty latencies
when deleting files. Enable the admin to configure the size down.
We will look into having a more sane default instead of UINT_MAX
sectors.
- Improvement of the SGP gaps logic from Keith Busch.
- Enable the block core to handle arbitrarily sized bios, which
enables a nice simplification of bio_add_page() (which is an IO hot
path). From Kent.
- Improvements to the partition io stats accounting, making it
faster. From Ming Lei.
- Also from Ming Lei, a basic fixup for overflow of the sysfs pending
file in blk-mq, as well as a fix for a blk-mq timeout race
condition.
- Ming Lin has been carrying Kents above mentioned patches forward
for a while, and testing them. Ming also did a few fixes around
that.
- Sasha Levin found and fixed a use-after-free problem introduced by
the bio->bi_error changes from Christoph.
- Small blk cgroup cleanup from Viresh Kumar"
* 'for-4.3/core' of git://git.kernel.dk/linux-block: (26 commits)
blk: Fix bio_io_vec index when checking bvec gaps
block: Replace SG_GAPS with new queue limits mask
block: bump BLK_DEF_MAX_SECTORS to 2560
Revert "block: remove artifical max_hw_sectors cap"
blk-mq: fix race between timeout and freeing request
blk-mq: fix buffer overflow when reading sysfs file of 'pending'
Documentation: update notes in biovecs about arbitrarily sized bios
block: remove bio_get_nr_vecs()
fs: use helper bio_add_page() instead of open coding on bi_io_vec
block: kill merge_bvec_fn() completely
md/raid5: get rid of bio_fits_rdev()
md/raid5: split bio for chunk_aligned_read
block: remove split code in blkdev_issue_{discard,write_same}
btrfs: remove bio splitting and merge_bvec_fn() calls
bcache: remove driver private bio splitting code
block: simplify bio_add_page()
block: make generic_make_request handle arbitrarily sized bios
blk-cgroup: Drop unlikely before IS_ERR(_OR_NULL)
block: don't access bio->bi_error after bio_put()
block: shrink struct bio down to 2 cache lines again
...
When a write to one of the legs of a RAID1 fails, the failure is
recorded in the metadata of the other leg(s) so that after a restart
the data on the failed drive wont be trusted even if that drive seems
to be working again (maybe a cable was unplugged).
Similarly when we record a bad-block in response to a write failure,
we must not let the write complete until the bad-block update is safe.
Currently there is no interlock between the write request completing
and the metadata update. So it is possible that the write will
complete, the app will confirm success in some way, and then the
machine will crash before the metadata update completes.
This is an extremely small hole for a racy to fit in, but it is
theoretically possible and so should be closed.
So:
- set MD_CHANGE_PENDING when requesting a metadata update for a
failed device, so we can know with certainty when it completes
- queue requests that experienced an error on a new queue which
is only processed after the metadata update completes
- call raid_end_bio_io() on bios in that queue when the time comes.
Signed-off-by: NeilBrown <neilb@suse.com>
When checking sync_action in a script, we want to be sure it is
as accurate as possible.
As resync/reshape etc doesn't always start immediately (a separate
thread is scheduled to do it), it is best if 'action_show'
checks if MD_RECOVER_NEEDED is set (which it does) and in that
case reports what is likely to start soon (which it only sometimes
does).
So:
- report 'reshape' if reshape_position suggests one might start.
- set MD_RECOVERY_RECOVER in raid1_reshape(), because that is very
likely to happen next.
Signed-off-by: NeilBrown <neilb@suse.com>
As generic_make_request() is now able to handle arbitrarily sized bios,
it's no longer necessary for each individual block driver to define its
own ->merge_bvec_fn() callback. Remove every invocation completely.
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Lars Ellenberg <drbd-dev@lists.linbit.com>
Cc: drbd-user@lists.linbit.com
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Yehuda Sadeh <yehuda@inktank.com>
Cc: Sage Weil <sage@inktank.com>
Cc: Alex Elder <elder@kernel.org>
Cc: ceph-devel@vger.kernel.org
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: dm-devel@redhat.com
Cc: Neil Brown <neilb@suse.de>
Cc: linux-raid@vger.kernel.org
Cc: Christoph Hellwig <hch@infradead.org>
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>
Acked-by: NeilBrown <neilb@suse.de> (for the 'md' bits)
Acked-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
[dpark: also remove ->merge_bvec_fn() in dm-thin as well as
dm-era-target, and resolve merge conflicts]
Signed-off-by: Dongsu Park <dpark@posteo.net>
Signed-off-by: Ming Lin <ming.l@ssi.samsung.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
raid1_end_read_request() assumes that the In_sync bits are consistent
with the ->degaded count.
raid1_spare_active updates the In_sync bit before the ->degraded count
and so exposes an inconsistency, as does error()
So extend the spinlock in raid1_spare_active() and error() to hide those
inconsistencies.
This should probably be part of
Commit: 34cab6f420 ("md/raid1: fix test for 'was read error from
last working device'.")
as it addresses the same issue. It fixes the same bug and should go
to -stable for same reasons.
Fixes: 76073054c9 ("md/raid1: clean up read_balance.")
Cc: stable@vger.kernel.org (v3.0+)
Signed-off-by: NeilBrown <neilb@suse.com>
Some places use helpers now, others don't. We only have the 'is set'
helper, add helpers for setting and clearing flags too.
It was a bit of a mess of atomic vs non-atomic access. With
BIO_UPTODATE gone, we don't have any risk of concurrent access to the
flags. So relax the restriction and don't make any of them atomic. The
flags that do have serialization issues (reffed and chained), we
already handle those separately.
Signed-off-by: Jens Axboe <axboe@fb.com>
Currently we have two different ways to signal an I/O error on a BIO:
(1) by clearing the BIO_UPTODATE flag
(2) by returning a Linux errno value to the bi_end_io callback
The first one has the drawback of only communicating a single possible
error (-EIO), and the second one has the drawback of not beeing persistent
when bios are queued up, and are not passed along from child to parent
bio in the ever more popular chaining scenario. Having both mechanisms
available has the additional drawback of utterly confusing driver authors
and introducing bugs where various I/O submitters only deal with one of
them, and the others have to add boilerplate code to deal with both kinds
of error returns.
So add a new bi_error field to store an errno value directly in struct
bio and remove the existing mechanisms to clean all this up.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
During a node failure, We need to suspend read balancing so that the
reads are directed to the first device and stale data is not read.
Suspending writes is not required because these would be recorded and
synced eventually.
A new flag MD_CLUSTER_SUSPEND_READ_BALANCING is set in recover_prep().
area_resyncing() will respond true for the entire devices if this
flag is set and the request type is READ. The flag is cleared
in recover_done().
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
Reported-By: David Teigland <teigland@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
When we get a read error from the last working device, we don't
try to repair it, and don't fail the device. We simple report a
read error to the caller.
However the current test for 'is this the last working device' is
wrong.
When there is only one fully working device, it assumes that a
non-faulty device is that device. However a spare which is rebuilding
would be non-faulty but so not the only working device.
So change the test from "!Faulty" to "In_sync". If ->degraded says
there is only one fully working device and this device is in_sync,
this must be the one.
This bug has existed since we allowed read_balance to read from
a recovering spare in v3.0
Reported-and-tested-by: Alexander Lyakas <alex.bolshoy@gmail.com>
Fixes: 76073054c9 ("md/raid1: clean up read_balance.")
Cc: stable@vger.kernel.org (v3.0+)
Signed-off-by: NeilBrown <neilb@suse.com>
Currently, a bdi (backing_dev_info) embeds single wb (bdi_writeback)
and the role of the separation is unclear. For cgroup support for
writeback IOs, a bdi will be updated to host multiple wb's where each
wb serves writeback IOs of a different cgroup on the bdi. To achieve
that, a wb should carry all states necessary for servicing writeback
IOs for a cgroup independently.
This patch moves bdi->state into wb.
* enum bdi_state is renamed to wb_state and the prefix of all enums is
changed from BDI_ to WB_.
* Explicit zeroing of bdi->state is removed without adding zeoring of
wb->state as the whole data structure is zeroed on init anyway.
* As there's still only one bdi_writeback per backing_dev_info, all
uses of bdi->state are mechanically replaced with bdi->wb.state
introducing no behavior changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: drbd-dev@lists.linbit.com
Cc: Neil Brown <neilb@suse.de>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This option is not well justified and testing suggests that
it hardly ever makes any difference.
The comment suggests there might be a need to wait for non-resync
activity indicated by ->nr_waiting, however raise_barrier()
already waits for all of that.
So just remove it to simplify reasoning about speed limiting.
This allows us to remove a 'FIXME' comment from raid5.c as that
never used the flag.
Signed-off-by: NeilBrown <neilb@suse.de>
When a drive is marked write-mostly it should only be the
target of reads if there is no other option.
This behaviour was broken by
commit 9dedf60313
md/raid1: read balance chooses idlest disk for SSD
which causes a write-mostly device to be *preferred* is some cases.
Restore correct behaviour by checking and setting
best_dist_disk and best_pending_disk rather than best_disk.
We only need to test one of these as they are both changed
from -1 or >=0 at the same time.
As we leave min_pending and best_dist unchanged, any non-write-mostly
device will appear better than the write-mostly device.
Reported-by: Tomáš Hodek <tomas.hodek@volny.cz>
Reported-by: Dark Penguin <darkpenguin@yandex.ru>
Signed-off-by: NeilBrown <neilb@suse.de>
Link: http://marc.info/?l=linux-raid&m=135982797322422
Fixes: 9dedf60313
Cc: stable@vger.kernel.org (3.6+)
Algorithm:
1. Node 1 issues mdadm --manage /dev/mdX --add /dev/sdYY which issues
ioctl(ADD_NEW_DISC with disc.state set to MD_DISK_CLUSTER_ADD)
2. Node 1 sends NEWDISK with uuid and slot number
3. Other nodes issue kobject_uevent_env with uuid and slot number
(Steps 4,5 could be a udev rule)
4. In userspace, the node searches for the disk, perhaps
using blkid -t SUB_UUID=""
5. Other nodes issue either of the following depending on whether the disk
was found:
ioctl(ADD_NEW_DISK with disc.state set to MD_DISK_CANDIDATE and
disc.number set to slot number)
ioctl(CLUSTERED_DISK_NACK)
6. Other nodes drop lock on no-new-devs (CR) if device is found
7. Node 1 attempts EX lock on no-new-devs
8. If node 1 gets the lock, it sends METADATA_UPDATED after unmarking the disk
as SpareLocal
9. If not (get no-new-dev lock), it fails the operation and sends METADATA_UPDATED
10. Other nodes understand if the device is added or not by reading the superblock again after receiving the METADATA_UPDATED message.
Signed-off-by: Lidong Zhong <lzhong@suse.com>
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
set choose_first true for cluster read in read balance when the area
is resyncing.
Signed-off-by: Lidong Zhong <lzhong@suse.com>
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
If there is a resync going on, all nodes must suspend writes to the
range. This is recorded in the suspend_info/suspend_list.
If there is an I/O within the ranges of any of the suspend_info,
should_suspend will return 1.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.com>
This modifies raid1's narrow_write_error to round up block_sectors to the
device's logical block size.
This prevents sd complaining about "Bad block number requested" for non-512-byte
sector disks.
Signed-off-by: Nate Dailey <nate.dailey@stratus.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Now that the ->stop function only frees the private data,
rename is accordingly.
Also pass in the private pointer as an arg rather than using
mddev->private. This flexibility will be useful in level_store().
Finally, don't clear ->private. It doesn't make sense to clear
it seeing that isn't what we free, and it is no longer necessary
to clear ->private (it was some time ago before ->to_remove was
introduced).
Setting ->to_remove in ->free() is a bit of a wart, but not a
big problem at the moment.
Signed-off-by: NeilBrown <neilb@suse.de>
Each md personality has a 'stop' operation which does two
things:
1/ it finalizes some aspects of the array to ensure nothing
is accessing the ->private data
2/ it frees the ->private data.
All the steps in '1' can apply to all arrays and so can be
performed in common code.
This is useful as in the case where we change the personality which
manages an array (in level_store()), it would be helpful to do
step 1 early, and step 2 later.
So split the 'step 1' functionality out into a new mddev_detach().
Signed-off-by: NeilBrown <neilb@suse.de>
There is no locking around calls to merge_bvec_fn(), so
it is possible that calls which coincide with a level (or personality)
change could go wrong.
So create a central dispatch point for these functions and use
rcu_read_lock().
If the array is suspended, reject any merge that can be rejected.
If not, we know it is safe to call the function.
Signed-off-by: NeilBrown <neilb@suse.de>
There is currently no locking around calls to the 'congested'
bdi function. If called at an awkward time while an array is
being converted from one level (or personality) to another, there
is a tiny chance of running code in an unreferenced module etc.
So add a 'congested' function to the md_personality operations
structure, and call it with appropriate locking from a central
'mddev_congested'.
When the array personality is changing the array will be 'suspended'
so no IO is processed.
If mddev_congested detects this, it simply reports that the
array is congested, which is a safe guess.
As mddev_suspend calls synchronize_rcu(), mddev_congested can
avoid races by included the whole call inside an rcu_read_lock()
region.
This require that the congested functions for all subordinate devices
can be run under rcu_lock. Fortunately this is the case.
Signed-off-by: NeilBrown <neilb@suse.de>
If a devices is being recovered it is not InSync and is not Faulty.
If a read error is experienced on that device, fix_read_error()
will be called, but it ignores non-InSync devices. So it will
neither fix the error nor fail the device.
It is incorrect that fix_read_error() ignores non-InSync devices.
It should only ignore Faulty devices. So fix it.
This became a bug when we allowed reading from a device that was being
recovered. It is suitable for any subsequent -stable kernel.
Fixes: da8840a747
Cc: stable@vger.kernel.org (v3.5+)
Reported-by: Alexander Lyakas <alex.bolshoy@gmail.com>
Tested-by: Alexander Lyakas <alex.bolshoy@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Both normal IO and resync IO can be retried with reschedule_retry()
and so be counted into ->nr_queued, but only normal IO gets counted in
->nr_pending.
Before the recent improvement to RAID1 resync there could only
possibly have been one or the other on the queue. When handling a
read failure it could only be normal IO. So when handle_read_error()
called freeze_array() the fact that freeze_array only compares
->nr_queued against ->nr_pending was safe.
But now that these two types can interleave, we can have both normal
and resync IO requests queued, so we need to count them both in
nr_pending.
This error can lead to freeze_array() hanging if there is a read
error, so it is suitable for -stable.
Fixes: 79ef3a8aa1
cc: stable@vger.kernel.org (v3.13+)
Reported-by: Brassow Jonathan <jbrassow@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.de>
raise_barrier() uses next_resync as part of its calculations, so it
really should be updated first, instead of afterwards.
next_resync is always used under resync_lock so update it under
resync lock to, just before it is used. That is safest.
This could cause normal IO and resync IO to interact badly so
it suitable for -stable.
Fixes: 79ef3a8aa1
cc: stable@vger.kernel.org (v3.13+)
Signed-off-by: NeilBrown <neilb@suse.de>
next_resync is (approximately) the location for the next resync request.
However it does *not* reliably determine the earliest location
at which resync might be happening.
This is because resync requests can complete out of order, and
we only limit the number of current requests, not the distance
from the earliest pending request to the latest.
mddev->curr_resync_completed is a reliable indicator of the earliest
position at which resync could be happening. It is updated less
frequently, but is actually reliable which is more important.
So use it to determine if a write request is before the region
being resynced and so safe from conflict.
This error can allow resync IO to interfere with normal IO which
could lead to data corruption. Hence: stable.
Fixes: 79ef3a8aa1
cc: stable@vger.kernel.org (v3.13+)
Signed-off-by: NeilBrown <neilb@suse.de>
