[ Upstream commit 33da8e7c81 ]
My recent to change to only use force_sig for a synchronous events
wound up breaking signal reception cifs and drbd. I had overlooked
the fact that by default kthreads start out with all signals set to
SIG_IGN. So a change I thought was safe turned out to have made it
impossible for those kernel thread to catch their signals.
Reverting the work on force_sig is a bad idea because what the code
was doing was very much a misuse of force_sig. As the way force_sig
ultimately allowed the signal to happen was to change the signal
handler to SIG_DFL. Which after the first signal will allow userspace
to send signals to these kernel threads. At least for
wake_ack_receiver in drbd that does not appear actively wrong.
So correct this problem by adding allow_kernel_signal that will allow
signals whose siginfo reports they were sent by the kernel through,
but will not allow userspace generated signals, and update cifs and
drbd to call allow_kernel_signal in an appropriate place so that their
thread can receive this signal.
Fixing things this way ensures that userspace won't be able to send
signals and cause problems, that it is clear which signals the
threads are expecting to receive, and it guarantees that nothing
else in the system will be affected.
This change was partly inspired by similar cifs and drbd patches that
added allow_signal.
Reported-by: ronnie sahlberg <ronniesahlberg@gmail.com>
Reported-by: Christoph Böhmwalder <christoph.boehmwalder@linbit.com>
Tested-by: Christoph Böhmwalder <christoph.boehmwalder@linbit.com>
Cc: Steve French <smfrench@gmail.com>
Cc: Philipp Reisner <philipp.reisner@linbit.com>
Cc: David Laight <David.Laight@ACULAB.COM>
Fixes: 247bc9470b ("cifs: fix rmmod regression in cifs.ko caused by force_sig changes")
Fixes: 72abe3bcf0 ("signal/cifs: Fix cifs_put_tcp_session to call send_sig instead of force_sig")
Fixes: fee109901f ("signal/drbd: Use send_sig not force_sig")
Fixes: 3cf5d076fb ("signal: Remove task parameter from force_sig")
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c38f03511 ]
print_st_err() is defined with its 4th argument taking an
'enum drbd_state_rv' but its prototype use an int for it.
Fix this by using 'enum drbd_state_rv' in the prototype too.
Signed-off-by: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Signed-off-by: Roland Kammerer <roland.kammerer@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f708bd08ec ]
"suspending" IO is overloaded.
It can mean "do not allow new requests" (obviously),
but it also may mean "must not complete pending IO",
for example while the fencing handlers do their arbitration.
When adjusting disk options, we suspend io (disallow new requests), then
wait for the activity-log to become unused (drain all IO completions),
and possibly replace it with a new activity log of different size.
If the other "suspend IO" aspect is active, pending IO completions won't
happen, and we would block forever (unkillable drbdsetup process).
Fix this by skipping the activity log adjustment if the "al-extents"
setting did not change. Also, in case it did change, fail early without
blocking if it looks like we would block forever.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fe43ed97bb ]
Multiple failure scenario:
a) all good
Connected Primary/Secondary UpToDate/UpToDate
b) lose disk on Primary,
Connected Primary/Secondary Diskless/UpToDate
c) continue to write to the device,
changes only make it to the Secondary storage.
d) lose disk on Secondary,
Connected Primary/Secondary Diskless/Diskless
e) now try to re-attach on Primary
This would have succeeded before, even though that is clearly the
wrong data set to attach to (missing the modifications from c).
Because we only compared our "effective" and the "to-be-attached"
data generation uuid tags if (device->state.conn < C_CONNECTED).
Fix: change that constraint to (device->state.pdsk != D_UP_TO_DATE)
compare the uuids, and reject the attach.
This patch also tries to improve the reverse scenario:
first lose Secondary, then Primary disk,
then try to attach the disk on Secondary.
Before this patch, the attach on the Secondary succeeds, but since commit
drbd: disconnect, if the wrong UUIDs are attached on a connected peer
the Primary will notice unsuitable data, and drop the connection hard.
Though unfortunately at a point in time during the handshake where
we cannot easily abort the attach on the peer without more
refactoring of the handshake.
We now reject any attach to "unsuitable" uuids,
as long as we can see a Primary role,
unless we already have access to "good" data.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 94c43a13b8 ]
During handshake, if we are diskless ourselves, we used to accept any size
presented by the peer.
Which could be zero if that peer was just brought up and connected
to us without having a disk attached first, in which case both
peers would just "flip" their volume sizes.
Now, even a diskless node will ignore "zero" sizes
presented by a diskless peer.
Also a currently Diskless Primary will refuse to shrink during handshake:
it may be frozen, and waiting for a "suitable" local disk or peer to
re-appear (on-no-data-accessible suspend-io). If the peer is smaller
than what we used to be, it is not suitable.
The logic for a diskless node during handshake is now supposed to be:
believe the peer, if
- I don't have a current size myself
- we agree on the size anyways
- I do have a current size, am Secondary, and he has the only disk
- I do have a current size, am Primary, and he has the only disk,
which is larger than my current size
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8e9c523016 ]
There are two callers of this function and they both unlock the mutex so
this ends up being a double unlock.
Fixes: 44ed167da7 ("drbd: rcu_read_lock() and rcu_dereference() for tconn->net_conf")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 77ce56e2bf ]
Building with clang and KASAN, we get a warning about an overly large
stack frame on 32-bit architectures:
drivers/block/drbd/drbd_receiver.c:921:31: error: stack frame size of 1280 bytes in function 'conn_connect'
[-Werror,-Wframe-larger-than=]
We already allocate other data dynamically in this function, so
just do the same for the shash descriptor, which makes up most of
this memory.
Link: https://lore.kernel.org/lkml/20190617132440.2721536-1-arnd@arndb.de/
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Roland Kammerer <roland.kammerer@linbit.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9848b6ddd8 ]
If you try to promote a Secondary while connected to a Primary
and allow-two-primaries is NOT set, we will wait for "ping-timeout"
to give this node a chance to detect a dead primary,
in case the cluster manager noticed faster than we did.
But if we then are *still* connected to a Primary,
we fail (after an additional timeout of ping-timout).
This change skips the spurious second timeout.
Most people won't notice really,
since "ping-timeout" by default is half a second.
But in some installations, ping-timeout may be 10 or 20 seconds or more,
and spuriously delaying the error return becomes annoying.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b17b59602b ]
With "on-no-data-accessible suspend-io", DRBD requires the next attach
or connect to be to the very same data generation uuid tag it lost last.
If we first lost connection to the peer,
then later lost connection to our own disk,
we would usually refuse to re-connect to the peer,
because it presents the wrong data set.
However, if the peer first connects without a disk,
and then attached its disk, we accepted that same wrong data set,
which would be "unexpected" by any user of that DRBD
and cause "undefined results" (read: very likely data corruption).
The fix is to forcefully disconnect as soon as we notice that the peer
attached to the "wrong" dataset.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d29e89e349 ]
So far there was the possibility that we called
genlmsg_new(GFP_NOIO)/mutex_lock() while holding an rcu_read_lock().
This included cases like:
drbd_sync_handshake (acquire the RCU lock)
drbd_asb_recover_1p
drbd_khelper
drbd_bcast_event
genlmsg_new(GFP_NOIO) --> may sleep
drbd_sync_handshake (acquire the RCU lock)
drbd_asb_recover_1p
drbd_khelper
notify_helper
genlmsg_new(GFP_NOIO) --> may sleep
drbd_sync_handshake (acquire the RCU lock)
drbd_asb_recover_1p
drbd_khelper
notify_helper
mutex_lock --> may sleep
While using GFP_ATOMIC whould have been possible in the first two cases,
the real fix is to narrow the rcu_read_lock.
Reported-by: Jia-Ju Bai <baijiaju1990@163.com>
Reviewed-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Roland Kammerer <roland.kammerer@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 64dafbc953 upstream.
We have
struct drbd_requests { ... struct bio *private_bio; ... }
to hold a bio clone for local submission.
On local IO completion, we put that bio, and in case we want to use the
result later, we overload that member to hold the ERR_PTR() of the
completion result,
Which, before v4.3, used to be the passed in "int error",
so we could first bio_put(), then assign.
v4.3-rc1~100^2~21 4246a0b63b block: add a bi_error field to struct bio
changed that:
bio_put(req->private_bio);
- req->private_bio = ERR_PTR(error);
+ req->private_bio = ERR_PTR(bio->bi_error);
Which introduces an access after free,
because it was non obvious that req->private_bio == bio.
Impact of that was mostly unnoticable, because we only use that value
in a multiple-failure case, and even then map any "unexpected" error
code to EIO, so worst case we could potentially mask a more specific
error with EIO in a multiple failure case.
Unless the pointed to memory region was unmapped, as is the case with
CONFIG_DEBUG_PAGEALLOC, in which case this results in
BUG: unable to handle kernel paging request
v4.13-rc1~70^2~75 4e4cbee93d block: switch bios to blk_status_t
changes it further to
bio_put(req->private_bio);
req->private_bio = ERR_PTR(blk_status_to_errno(bio->bi_status));
And blk_status_to_errno() now contains a WARN_ON_ONCE() for unexpected
values, which catches this "sometimes", if the memory has been reused
quickly enough for other things.
Should also go into stable since 4.3, with the trivial change around 4.13.
Cc: stable@vger.kernel.org
Fixes: 4246a0b63b block: add a bi_error field to struct bio
Reported-by: Sarah Newman <srn@prgmr.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Don't pass a size larger than iov_len to kernel_sendmsg().
Otherwise it will cause a NULL pointer deref when kernel_sendmsg()
returns with rv < size.
DRBD as external module has been around in the kernel 2.4 days already.
We used to be compatible to 2.4 and very early 2.6 kernels,
we used to use
rv = sock_sendmsg(sock, &msg, iov.iov_len);
then later changed to
rv = kernel_sendmsg(sock, &msg, &iov, 1, size);
when we should have used
rv = kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len);
tcp_sendmsg() used to totally ignore the size parameter.
57be5bd ip: convert tcp_sendmsg() to iov_iter primitives
changes that, and exposes our long standing error.
Even with this error exposed, to trigger the bug, we would need to have
an environment (config or otherwise) causing us to not use sendpage()
for larger transfers, a failing connection, and have it fail "just at the
right time". Apparently that was unlikely enough for most, so this went
unnoticed for years.
Still, it is known to trigger at least some of these,
and suspected for the others:
[0] http://lists.linbit.com/pipermail/drbd-user/2016-July/023112.html
[1] http://lists.linbit.com/pipermail/drbd-dev/2016-March/003362.html
[2] https://forums.grsecurity.net/viewtopic.php?f=3&t=4546
[3] https://ubuntuforums.org/showthread.php?t=2336150
[4] http://e2.howsolveproblem.com/i/1175162/
This should go into 4.9,
and into all stable branches since and including v4.0,
which is the first to contain the exposing change.
It is correct for all stable branches older than that as well
(which contain the DRBD driver; which is 2.6.33 and up).
It requires a small "conflict" resolution for v4.4 and earlier, with v4.5
we dropped the comment block immediately preceding the kernel_sendmsg().
Fixes: b411b3637f ("The DRBD driver")
Cc: <stable@vger.kernel.org> # 2.6.33.x-
Cc: viro@zeniv.linux.org.uk
Cc: christoph.lechleitner@iteg.at
Cc: wolfgang.glas@iteg.at
Reported-by: Christoph Lechleitner <christoph.lechleitner@iteg.at>
Tested-by: Christoph Lechleitner <christoph.lechleitner@iteg.at>
Signed-off-by: Richard Weinberger <richard@nod.at>
[changed oneliner to be "obvious" without context; more verbose message]
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.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>
Pull vfs updates from Al Viro:
"Assorted cleanups and fixes.
Probably the most interesting part long-term is ->d_init() - that will
have a bunch of followups in (at least) ceph and lustre, but we'll
need to sort the barrier-related rules before it can get used for
really non-trivial stuff.
Another fun thing is the merge of ->d_iput() callers (dentry_iput()
and dentry_unlink_inode()) and a bunch of ->d_compare() ones (all
except the one in __d_lookup_lru())"
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (26 commits)
fs/dcache.c: avoid soft-lockup in dput()
vfs: new d_init method
vfs: Update lookup_dcache() comment
bdev: get rid of ->bd_inodes
Remove last traces of ->sync_page
new helper: d_same_name()
dentry_cmp(): use lockless_dereference() instead of smp_read_barrier_depends()
vfs: clean up documentation
vfs: document ->d_real()
vfs: merge .d_select_inode() into .d_real()
unify dentry_iput() and dentry_unlink_inode()
binfmt_misc: ->s_root is not going anywhere
drop redundant ->owner initializations
ufs: get rid of redundant checks
orangefs: constify inode_operations
missed comment updates from ->direct_IO() prototype change
file_inode(f)->i_mapping is f->f_mapping
trim fsnotify hooks a bit
9p: new helper - v9fs_parent_fid()
debugfs: ->d_parent is never NULL or negative
...
Pull block driver updates from Jens Axboe:
"This branch also contains core changes. I've come to the conclusion
that from 4.9 and forward, I'll be doing just a single branch. We
often have dependencies between core and drivers, and it's hard to
always split them up appropriately without pulling core into drivers
when that happens.
That said, this contains:
- separate secure erase type for the core block layer, from
Christoph.
- set of discard fixes, from Christoph.
- bio shrinking fixes from Christoph, as a followup up to the
op/flags change in the core branch.
- map and append request fixes from Christoph.
- NVMeF (NVMe over Fabrics) code from Christoph. This is pretty
exciting!
- nvme-loop fixes from Arnd.
- removal of ->driverfs_dev from Dan, after providing a
device_add_disk() helper.
- bcache fixes from Bhaktipriya and Yijing.
- cdrom subchannel read fix from Vchannaiah.
- set of lightnvm updates from Wenwei, Matias, Johannes, and Javier.
- set of drbd updates and fixes from Fabian, Lars, and Philipp.
- mg_disk error path fix from Bart.
- user notification for failed device add for loop, from Minfei.
- NVMe in general:
+ NVMe delay quirk from Guilherme.
+ SR-IOV support and command retry limits from Keith.
+ fix for memory-less NUMA node from Masayoshi.
+ use UINT_MAX for discard sectors, from Minfei.
+ cancel IO fixes from Ming.
+ don't allocate unused major, from Neil.
+ error code fixup from Dan.
+ use constants for PSDT/FUSE from James.
+ variable init fix from Jay.
+ fabrics fixes from Ming, Sagi, and Wei.
+ various fixes"
* 'for-4.8/drivers' of git://git.kernel.dk/linux-block: (115 commits)
nvme/pci: Provide SR-IOV support
nvme: initialize variable before logical OR'ing it
block: unexport various bio mapping helpers
scsi/osd: open code blk_make_request
target: stop using blk_make_request
block: simplify and export blk_rq_append_bio
block: ensure bios return from blk_get_request are properly initialized
virtio_blk: use blk_rq_map_kern
memstick: don't allow REQ_TYPE_BLOCK_PC requests
block: shrink bio size again
block: simplify and cleanup bvec pool handling
block: get rid of bio_rw and READA
block: don't ignore -EOPNOTSUPP blkdev_issue_write_same
block: introduce BLKDEV_DISCARD_ZERO to fix zeroout
NVMe: don't allocate unused nvme_major
nvme: avoid crashes when node 0 is memoryless node.
nvme: Limit command retries
loop: Make user notify for adding loop device failed
nvme-loop: fix nvme-loop Kconfig dependencies
nvmet: fix return value check in nvmet_subsys_alloc()
...
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>
crypto_alloc_hash returns an ERR_PTR(), not NULL.
Also reset peer_integrity_tfm to NULL, to not call crypto_free_hash()
on an errno in the cleanup path.
Reported-by: Insu Yun <wuninsu@gmail.com>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
For larger devices, the array of bitmap page pointers can grow very
large (8000 pointers per TB of storage).
For each activity log transaction, we need to flush the associated
bitmap pages to stable storage. Currently, we just "mark" the respective
pages while setting up the transaction, then tell the bitmap code to
write out all marked pages, but skip unchanged pages.
But one such transaction can affect only a small number of bitmap pages,
there is no need to scan the full array of several (ten-)thousand
page pointers to find the few marked ones.
Instead, remember the index numbers of the few affected pages,
and later only re-check those to skip duplicates and unchanged ones.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This should silence a warning about an empty statement. Thanks to Fabian
Frederick <fabf@skynet.be> who sent a patch I modified to be smaller and
avoids an additional indent level.
Signed-off-by: Roland Kammerer <roland.kammerer@linbit.com>
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This contains various cosmetic fixes ranging from simple typos to
const-ifying, and using booleans properly.
Original commit messages from Fabian's patch set:
drbd: debugfs: constify drbd_version_fops
drbd: use seq_put instead of seq_print where possible
drbd: include linux/uaccess.h instead of asm/uaccess.h
drbd: use const char * const for drbd strings
drbd: kerneldoc warning fix in w_e_end_data_req()
drbd: use unsigned for one bit fields
drbd: use bool for peer is_ states
drbd: fix typo
drbd: use | for bitmask combination
drbd: use true/false for bool
drbd: fix drbd_bm_init() comments
drbd: introduce peer state union
drbd: fix maybe_pull_ahead() locking comments
drbd: use bool for growing
drbd: remove redundant declarations
drbd: replace if/BUG by BUG_ON
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Roland Kammerer <roland.kammerer@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Scenario, starting with normal operation
Connected Primary/Secondary UpToDate/UpToDate
NetworkFailure Primary/Unknown UpToDate/DUnknown (frozen)
... more failures happen, secondary loses it's disk,
but eventually is able to re-establish the replication link ...
Connected Primary/Secondary UpToDate/Diskless (resumed; needs to bump uuid!)
We used to just resume/resent suspended requests,
without bumping the UUID.
Which will lead to problems later, when we want to re-attach the disk on
the peer, without first disconnecting, or if we experience additional
failures, because we now have diverging data without being able to
recognize it.
Make sure we also bump the current data generation UUID,
if we notice "peer disk unknown" -> "peer disk known bad".
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
We already serialize connection state changes,
and other, non-connection state changes (role changes)
while we are establishing a connection.
But if we have an established connection,
then trigger a resync handshake (by primary --force or similar),
until now we just had to be "lucky".
Consider this sequence (e.g. deployment scenario):
create-md; up;
-> Connected Secondary/Secondary Inconsistent/Inconsistent
then do a racy primary --force on both peers.
block drbd0: drbd_sync_handshake:
block drbd0: self 0000000000000004:0000000000000000:0000000000000000:0000000000000000 bits:25590 flags:0
block drbd0: peer 0000000000000004:0000000000000000:0000000000000000:0000000000000000 bits:25590 flags:0
block drbd0: peer( Unknown -> Secondary ) conn( WFReportParams -> Connected ) pdsk( DUnknown -> Inconsistent )
block drbd0: peer( Secondary -> Primary ) pdsk( Inconsistent -> UpToDate )
*** HERE things go wrong. ***
block drbd0: role( Secondary -> Primary )
block drbd0: drbd_sync_handshake:
block drbd0: self 0000000000000005:0000000000000000:0000000000000000:0000000000000000 bits:25590 flags:0
block drbd0: peer C90D2FC716D232AB:0000000000000004:0000000000000000:0000000000000000 bits:25590 flags:0
block drbd0: Becoming sync target due to disk states.
block drbd0: Writing the whole bitmap, full sync required after drbd_sync_handshake.
block drbd0: Remote failed to finish a request within 6007ms > ko-count (2) * timeout (30 * 0.1s)
drbd s0: peer( Primary -> Unknown ) conn( Connected -> Timeout ) pdsk( UpToDate -> DUnknown )
The problem here is that the local promotion happens before the sync handshake
triggered by the remote promotion was completed. Some assumptions elsewhere
become wrong, and when the expected resync handshake is then received and
processed, we get stuck in a deadlock, which can only be recovered by reboot :-(
Fix: if we know the peer has good data,
and our own disk is present, but NOT good,
and there is no resync going on yet,
we expect a sync handshake to happen "soon".
So reject a racy promotion with SS_IN_TRANSIENT_STATE.
Result:
... as above ...
block drbd0: peer( Secondary -> Primary ) pdsk( Inconsistent -> UpToDate )
*** local promotion being postponed until ... ***
block drbd0: drbd_sync_handshake:
block drbd0: self 0000000000000004:0000000000000000:0000000000000000:0000000000000000 bits:25590 flags:0
block drbd0: peer 77868BDA836E12A5:0000000000000004:0000000000000000:0000000000000000 bits:25590 flags:0
...
block drbd0: conn( WFBitMapT -> WFSyncUUID )
block drbd0: updated sync uuid 85D06D0E8887AD44:0000000000000000:0000000000000000:0000000000000000
block drbd0: conn( WFSyncUUID -> SyncTarget )
*** ... after the resync handshake ***
block drbd0: role( Secondary -> Primary )
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If in a two-primary scenario, we lost our peer, freeze IO,
and are still frozen (no UUID rotation) when the peer comes back
as Secondary after a hard crash, we will see identical UUIDs.
The "rule_nr = 40" chose to use the "CRASHED_PRIMARY" bit as
arbitration, but that would cause the still running (but frozen) Primary
to become SyncTarget (which it typically refuses), and the handshake is
declined.
Fix: check current roles.
If we have *one* current primary, the Primary wins.
(rule_nr = 41)
Since that is a protocol change, use the newly introduced DRBD_FF_WSAME
to determine if rule_nr = 41 can be applied.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
We will support WRITE_SAME, if
* all peers support WRITE_SAME (both in kernel and DRBD version),
* all peer devices support WRITE_SAME
* logical_block_size is identical on all peers.
We may at some point introduce a fallback on the receiving side
for devices/kernels that do not support WRITE_SAME,
by open-coding a submit loop. But not yet.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Even if discard_zeroes_data != 0,
if discard_zeroes_if_aligned is set, we assume we can reliably
zero-out/discard using the drbd_issue_peer_discard() helper.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
When re-attaching the local backend device to a C_STANDALONE D_DISKLESS
R_PRIMARY with OND_SUSPEND_IO, we may only resume IO if we recognize the
backend that is being attached as D_UP_TO_DATE.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If DRBD lost all path to good data,
and the on-no-data-accessible policy is OND_SUSPEND_IO,
all pending and new IO requests are suspended (will block).
If that setting is OND_IO_ERROR, IO will still be completed.
READ to "clean" areas (e.g. on an D_INCONSISTENT device,
and bitmap indicates a block is already in sync) will succeed.
READ to "unclean" areas (bitmap indicates block is out-of-sync),
will return EIO.
If we are already D_DISKLESS (or D_FAILED), we also return EIO.
Unfortunately, on a former R_PRIMARY C_SYNC_TARGET D_INCONSISTENT,
after replication link loss, new WRITE requests still went through OK.
The would also set the "out-of-sync" bit on their way, so READ after
WRITE would still return EIO. Also, the data generation UUIDs had not
been bumped, we would cause data divergence, without being able to
detect it on the next sync handshake, given the right sequence of events
in a multiple error scenario and "improper" order of recovery actions.
The right thing to do is to return EIO for all new writes,
unless we have access to good, current, D_UP_TO_DATE data.
The "established best practices" way to avoid these situations in the
first place is to set OND_SUSPEND_IO, or even do a hard-reset from
the pri-on-incon-degr policy helper hook.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Possibly sequence of events:
SyncTarget is made Primary, then loses replication link
(only path to good data on SyncSource).
Behavior is then controlled by the on-no-data-accessible policy,
which defaults to OND_IO_ERROR (may be set to OND_SUSPEND_IO).
If OND_IO_ERROR is in fact the current policy, we clear the susp_fen
(IO suspended due to fencing policy) flag, do NOT set the susp_nod
(IO suspended due to no data) flag.
But we forgot to call the IO error completion for all pending,
suspended, requests.
While at it, also add a race check for a theoretically possible
race with a new handshake (network hickup), we may be able to
re-send requests, and can avoid passing IO errors up the stack.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
When resync is finished, we already call the "after-resync-target"
handler (on the former sync target, obviously), once per volume.
Paired with the before-resync-target handler, you can create snapshots,
before the resync causes the volumes to become inconsistent,
and discard those snapshots again, once they are no longer needed.
It was also overloaded to be paired with the "fence-peer" handler,
to "unfence" once the volumes are up-to-date and known good.
This has some disadvantages, though: we call "fence-peer" for the whole
connection (once for the group of volumes), but would call unfence as
side-effect of after-resync-target once for each volume.
Also, we fence on a (current, or about to become) Primary,
which will later become the sync-source.
Calling unfence only as a side effect of the after-resync-target
handler opens a race window, between a new fence on the Primary
(SyncTarget) and the unfence on the SyncTarget, which is difficult to
close without some kind of "cluster wide lock" in those handlers.
We would not need those handlers if we could still communicate.
Which makes trying to aquire a cluster wide lock from those handlers
seem like a very bad idea.
This introduces the "unfence-peer" handler, which will be called
per connection (once for the group of volumes), just like the fence
handler, only once all volumes are back in sync, and on the SyncSource.
Which is expected to be the node that previously called "fence", the
node that is currently allowed to be Primary, and thus the only node
that could trigger a new "fence" that could race with this unfence.
Which makes us not need any cluster wide synchronization here,
serializing two scripts running on the same node is trivial.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If the replication link breaks exactly during "resync finished" detection,
finishing too early on the sync source could again lead to UUIDs rotated
too fast, and potentially a spurious full resync on next handshake.
Always wait for explicit resync finished state change notification from
the sync target.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Make sure we have at least 67 (> AL_UPDATES_PER_TRANSACTION)
al-extents available, and allow up to half of that to be
discarded in one bio.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
For consistency, also zero-out partial unaligned chunks of discard
requests on the local backend.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Now that we have the discard_zeroes_if_aligned setting, we should also
check it when setting up our queue parameters on the primary,
not only on the receiving side.
We announce discard support,
UNLESS
* we are connected to a peer that does not support TRIM
on the DRBD protocol level. Otherwise, it would either discard, or
do a fallback to zero-out, depending on its backend and configuration.
* our local backend does not support discards,
or (discard_zeroes_data=0 AND discard_zeroes_if_aligned=no).
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
We can avoid spurious data divergence caused by partially-ignored
discards on certain backends with discard_zeroes_data=0, if we
translate partial unaligned discard requests into explicit zero-out.
The relevant use case is LVM/DM thin.
If on different nodes, DRBD is backed by devices with differing
discard characteristics, discards may lead to data divergence
(old data or garbage left over on one backend, zeroes due to
unmapped areas on the other backend). Online verify would now
potentially report tons of spurious differences.
While probably harmless for most use cases (fstrim on a file system),
DRBD cannot have that, it would violate our promise to upper layers
that our data instances on the nodes are identical.
To be correct and play safe (make sure data is identical on both copies),
we would have to disable discard support, if our local backend (on a
Primary) does not support "discard_zeroes_data=true".
We'd also have to translate discards to explicit zero-out on the
receiving (typically: Secondary) side, unless the receiving side
supports "discard_zeroes_data=true".
Which both would allocate those blocks, instead of unmapping them,
in contrast with expectations.
LVM/DM thin does set discard_zeroes_data=0,
because it silently ignores discards to partial chunks.
We can work around this by checking the alignment first.
For unaligned (wrt. alignment and granularity) or too small discards,
we zero-out the initial (and/or) trailing unaligned partial chunks,
but discard all the aligned full chunks.
At least for LVM/DM thin, the result is effectively "discard_zeroes_data=1".
Arguably it should behave this way internally, by default,
and we'll try to make that happen.
But our workaround is still valid for already deployed setups,
and for other devices that may behave this way.
Setting discard-zeroes-if-aligned=yes will allow DRBD to use
discards, and to announce discard_zeroes_data=true, even on
backends that announce discard_zeroes_data=false.
Setting discard-zeroes-if-aligned=no will cause DRBD to always
fall-back to zero-out on the receiving side, and to not even
announce discard capabilities on the Primary, if the respective
backend announces discard_zeroes_data=false.
We used to ignore the discard_zeroes_data setting completely.
To not break established and expected behaviour, and suddenly
cause fstrim on thin-provisioned LVs to run out-of-space,
instead of freeing up space, the default value is "yes".
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
To maintain write-order fidelity accros all volumes in a DRBD resource,
the receiver of a P_BARRIER needs to issue flushes to all volumes.
We used to do this by calling blkdev_issue_flush(), synchronously,
one volume at a time.
We now submit all flushes to all volumes in parallel, then wait for all
completions, to reduce worst-case latencies on multi-volume resources.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
The command line parameter the kernel module uses to communicate the
device minor to userland helper is flawed in a way that the device
indentifier "minor-%d" is being truncated to minors with a maximum
of 5 digits.
But DRBD 8.4 allows 2^20 == 1048576 minors,
thus a minimum of 7 digits must be supported.
Reported by Veit Wahlich on drbd-dev.
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Regression introduced with 8.4.5
drbd: application writes may set-in-sync in protocol != C
Overwriting the same block (LBA) while a former version is still
"in-flight" to the peer (to be exact: we did not receive the
P_BARRIER_ACK for its epoch yet) would wait for the full epoch of that
former version to be acknowledged by the peer.
In synchronous and quasi-synchronous protocols C and B,
this may double the latency on overwrites.
With protocol A, which is supposed to be asynchronous and only wait for
local completion, it is even worse: it would make overwrites
quasi-synchronous, they would be hit by the full RTT, which protocol A
was specifically meant to avoid, and possibly the additional time it
takes to drain the buffers first.
Particularly bad for databases, or anything else that
does frequent updates to the same blocks (various file system meta data).
No impact if >= rtt passes between updates to the same block.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If thinly provisioned volumes are used, during a resync the sync source
tries to find out if a block is deallocated. If it is deallocated, then
the resync target uses block_dev_issue_zeroout() on the range in
question.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
As long as the value is 0 the feature is disabled. With setting
it to a positive value, DRBD limits and aligns its resync requests
to the rs-discard-granularity setting. If the sync source detects
all zeros in such a block, the resync target discards the range
on disk.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If during resync we read only zeroes for a range of sectors assume
that these secotors can be discarded on the sync target node.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
When leaving resync states because of disconnect,
do the bitmap write-out synchronously in the drbd_disconnected() path.
When leaving resync states because we go back to AHEAD/BEHIND, or
because resync actually finished, or some disk was lost during resync,
trigger the write-out from after_state_ch().
The bitmap write-out for resync -> ahead/behind was missing completely before.
Note that this is all only an optimization to avoid double-resyncs of
already completed blocks in case this node crashes.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
The intention was to only suspend IO if some normal bitmap operation is
supposed to be locked out, not always. If the bulk operation is flaged
as BM_LOCKED_CHANGE_ALLOWED, we do not need to suspend IO.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Use BIO_MAX_PAGES instead and we will remove BIO_MAX_SIZE.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Tested-by: Hannes Reinecke <hare@suse.com>
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>
