linux-stable.git/drivers/block/drbd/drbd_req.c, branch linux-4.19.y

drbd: fix potential silent data corruption

2022-04-15T12:14:40+00:00

commit f4329d1f848ac35757d9cc5487669d19dfc5979c upstream.

Scenario:
---------

bio chain generated by blk_queue_split().
Some split bio fails and propagates its error status to the "parent" bio.
But then the (last part of the) parent bio itself completes without error.

We would clobber the already recorded error status with BLK_STS_OK,
causing silent data corruption.

Reproducer:
-----------

How to trigger this in the real world within seconds:

DRBD on top of degraded parity raid,
small stripe_cache_size, large read_ahead setting.
Drop page cache (sysctl vm.drop_caches=1, fadvise "DONTNEED",
umount and mount again, "reboot").

Cause significant read ahead.

Large read ahead request is split by blk_queue_split().
Parts of the read ahead that are already in the stripe cache,
or find an available stripe cache to use, can be serviced.
Parts of the read ahead that would need "too much work",
would need to wait for a "stripe_head" to become available,
are rejected immediately.

For larger read ahead requests that are split in many pieces, it is very
likely that some "splits" will be serviced, but then the stripe cache is
exhausted/busy, and the remaining ones will be rejected.

Signed-off-by: Lars Ellenberg 
Signed-off-by: Christoph Böhmwalder 
Cc:  # 4.13.x
Link: https://lore.kernel.org/r/20220330185551.3553196-1-christoph.boehmwalder@linbit.com
Signed-off-by: Jens Axboe 
Signed-off-by: Greg Kroah-Hartman

block: Add and use op_stat_group() for indexing disk_stat fields.

2018-07-18T14:44:20+00:00

Add and use a new op_stat_group() function for indexing partition stat
fields rather than indexing them by rq_data_dir() or bio_data_dir().
This function works similarly to op_is_sync() in that it takes the
request::cmd_flags or bio::bi_opf flags and determines which stats
should et updated.

In addition, the second parameter to generic_start_io_acct() and
generic_end_io_acct() is now a REQ_OP rather than simply a read or
write bit and it uses op_stat_group() on the parameter to determine
the stat group.

Note that the partition in_flight counts are not part of the per-cpu
statistics and as such are not indexed via this function.  It's now
indexed by op_is_write().

tj: Refreshed on top of v4.17.  Updated to pass around REQ_OP.

Signed-off-by: Michael Callahan 
Signed-off-by: Tejun Heo 
Cc: Minchan Kim 
Cc: Dan Williams 
Cc: Joshua Morris 
Cc: Philipp Reisner 
Cc: Matias Bjorling 
Cc: Kent Overstreet 
Cc: Alasdair Kergon 
Signed-off-by: Jens Axboe

drbd: Fix drbd_request_prepare() discard handling

2018-06-29T13:53:21+00:00

Fix the test that verifies whether bio_op(bio) represents a discard
or write zeroes operation. Compile-tested only.

Cc: Philipp Reisner 
Cc: Lars Ellenberg 
Fixes: 7435e9018f91 ("drbd: zero-out partial unaligned discards on local backend")
Signed-off-by: Bart Van Assche 
Reviewed-by: Christoph Hellwig 
Signed-off-by: Jens Axboe

drbd: convert to bioset_init()/mempool_init()

2018-05-30T21:33:32+00:00

Convert drbd to embedded bio sets and mempools.

Signed-off-by: Kent Overstreet 
Signed-off-by: Jens Axboe

drbd: Convert timers to use timer_setup()

2017-11-06T20:49:57+00:00

In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly.

Cc: Philipp Reisner 
Cc: Lars Ellenberg 
Cc: drbd-dev@lists.linbit.com
Signed-off-by: Kees Cook

drbd: add explicit plugging when submitting batches

2017-08-29T21:34:44+00:00

When submitting batches of requests which had been queued on the
submitter thread, typically because they needed to wait for an
activity log transactions, use explicit plugging to help potential
merging of requests in the backend io-scheduler.

Signed-off-by: Philipp Reisner 
Signed-off-by: Lars Ellenberg 
Signed-off-by: Jens Axboe

drbd: change list_for_each_safe to while(list_first_entry_or_null)

2017-08-29T21:34:44+00:00

Two instances of list_for_each_safe can drop their tmp element, they
really just peel off each element in turn from the start of the list.

Signed-off-by: Philipp Reisner 
Signed-off-by: Lars Ellenberg 
Signed-off-by: Jens Axboe

drbd: introduce drbd_recv_header_maybe_unplug

2017-08-29T21:34:43+00:00

Recently, drbd_recv_header() was changed to potentially
implicitly "unplug" the backend device(s), in case there
is currently nothing to receive.

Be more explicit about it: re-introduce the original drbd_recv_header(),
and introduce a new drbd_recv_header_maybe_unplug() for use by the
receiver "main loop".

Using explicit plugging via blk_start_plug(); blk_finish_plug();
really helps the io-scheduler of the backend with merging requests.

Wrap the receiver "main loop" with such a plug.
Also catch unplug events on the Primary,
and try to propagate.

This is performance relevant.  Without this, if the receiving side does
not merge requests, number of IOPS on the peer can me significantly
higher than IOPS on the Primary, and can easily become the bottleneck.

Together, both changes should help to reduce the number of IOPS
as seen on the backend of the receiving side, by increasing
the chance of merging mergable requests, without trading latency
for more throughput.

Signed-off-by: Philipp Reisner 
Signed-off-by: Lars Ellenberg 
Signed-off-by: Jens Axboe

block: replace bi_bdev with a gendisk pointer and partitions index

2017-08-23T18:49:55+00:00

This way we don't need a block_device structure to submit I/O.  The
block_device has different life time rules from the gendisk and
request_queue and is usually only available when the block device node
is open.  Other callers need to explicitly create one (e.g. the lightnvm
passthrough code, or the new nvme multipathing code).

For the actual I/O path all that we need is the gendisk, which exists
once per block device.  But given that the block layer also does
partition remapping we additionally need a partition index, which is
used for said remapping in generic_make_request.

Note that all the block drivers generally want request_queue or
sometimes the gendisk, so this removes a layer of indirection all
over the stack.

Signed-off-by: Christoph Hellwig 
Signed-off-by: Jens Axboe

block: pass in queue to inflight accounting

2017-08-09T19:09:16+00:00

No functional change in this patch, just in preparation for
basing the inflight mechanism on the queue in question.

Reviewed-by: Bart Van Assche 
Reviewed-by: Omar Sandoval 
Signed-off-by: Jens Axboe