The existing ioctl interface for configuring NBD devices is a bit
cumbersome and hard to extend.  The other problem is we leave a
userspace app sitting in it's syscall until the device disconnects,
which is less than ideal.

This patch introduces a netlink interface for adding and disconnecting
nbd devices.  This has the benefits of being easily extendable without
breaking older userspace applications, and allows us to configure a nbd
device without leaving a userspace app sitting waiting for the device to
disconnect.

With this interface we also gain the ability to configure more devices
than are preallocated at insmod time.  We also have gained the ability
to not specify a particular device and be provided one for us so that
userspace doesn't need to find a free device to configure.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

In preparation for the upcoming netlink interface we need to not rely on
already having the bdev for the NBD device we are doing operations on.
Instead of passing the bdev around, just use it in places where we know
we already have the bdev.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

In order to properly refcount the various aspects of a NBD device we
need to separate out the configuration elements of the nbd device.  The
configuration of a NBD device has a different lifetime from the actual
device, so it doesn't make sense to bundle these two concepts.  Add a
config_refs to keep track of the configuration structure, that way we
can be sure that we never access it when we've torn down the device.
Add a new nbd_config structure to hold all of the transient
configuration information.  Finally create this when we open the device
so that it is in place when we start to configure the device.  This has
a nice side-effect of fixing a long standing problem where you could end
up with a half-configured nbd device that needed to be "disconnected" in
order to be usable again.  Now once we close our device the
configuration will be discarded.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently if we have multiple connections and one of them goes down we will tear
down the whole device.  However there's no reason we need to do this as we
could have other connections that are working fine.  Deal with this by keeping
track of the state of the different connections, and if we lose one we mark it
as dead and send all IO destined for that socket to one of the other healthy
sockets.  Any outstanding requests that were on the dead socket will timeout and
be re-submitted properly.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When adding a new socket we look it up and then try to add it to our
configuration.  If any of those steps fail we need to make sure we put
the socket so we don't leak them.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Now that we use the proper REQ_OP_WRITE_ZEROES operation everywhere we can
kill this hack.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Constify all instances of blk_mq_ops, as they are never modified.
Signed-off-by: NEric Biggers <ebiggers@google.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When a filesystem is mounted on a nbd device and on a disconnect, because
of kill_bdev(), and resetting bdev size to zero, buffer_head mappings are
getting destroyed under mounted filesystem.

After a bdev size reset(i.e bdev->bd_inode->i_size = 0) on a disconnect,
followed by a sys_umount(),
        generic_shutdown_super()->...
        ->__sync_blockdev()->...
        -blkdev_writepages()->...
        ->do_invalidatepage()->...
        -discard_buffer()   is discarding superblock buffer_head assumed
to be in mapped state by ext4_commit_super().

[mlin: ported to 4.11-rc2]
Signed-off-by: Ratna Manoj Bolla <manoj.br@gmail.com
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We can't just set the timeout on the tagset, we have to set it on the
queue as it would have been setup already at this point.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We've been relying on the block layer to assume rq->errors being set
translates into -EIO.  I noticed in testing that sometimes this isn't
true, and really there's not much of a reason to have a counter instead
of just using -EIO.  So set it properly so we don't leak random numbers
to unsuspecting victims.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We can submit IO in a processes context, which means there can be
pending signals.  This isn't a fatal error for NBD, but it does require
some finesse.  If the signal happens before we transmit anything then we
are ok, just requeue the request and carry on.  However if we've done a
partial transmit we can't allow anything else to be transmitted on this
socket until we transmit the remaining part of the request.  Deal with
this by keeping track of how much we've sent for the current request,
and if we get an ERESTARTSYS during any part of our transmission save
the state of that request and requeue the IO.  If anybody tries to
submit a request that isn't our pending request then requeue that
request until we are able to service the one that is pending.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This was introduced in the multi-connection patch, we've been leaking
socket's ever since.

Fixes: 9561a7ad ("nbd: add multi-connection support")
cc: stable@vger.kernel.org
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

If we fail to register the blockdev we need to make sure to destroy the
recv workqueue.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We noticed when trying to do O_DIRECT to an export on the server side
that we were getting requests smaller than the 4k sectorsize of the
device.  This is because the client isn't setting the logical and
physical blocksizes properly for the underlying device.  Fix this up by
setting the queue blocksizes and then calling bd_set_size.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Break the ioctl handling out into helper functions, some of these things
are getting pretty big and unwieldy.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

To prepare for dynamically adding new nbd devices to the system switch
from using an array for the nbd devices and instead use an idr.  This
copies what loop does for keeping track of its devices.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Since we are in the memory reclaim path we need our recv work to be on a
workqueue that has WQ_MEM_RECLAIM set so we can avoid deadlocks.  Also
set WQ_HIGHPRI since we are in the completion path for IO.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Instead of keeping two levels of indirection for requests types, fold it
all into the operations.  The little caveat here is that previously
cmd_type only applied to struct request, while the request and bio op
fields were set to plain REQ_OP_READ/WRITE even for passthrough
operations.

Instead this patch adds new REQ_OP_* for SCSI passthrough and driver
private requests, althought it has to add two for each so that we
can communicate the data in/out nature of the request.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

This is where we do the rest of the request handling, which will
become much simpler soon, too.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

Disconnects don't use block layer requests these days, so all handling
of private requests is dead code.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

A user noticed that write performance was horrible over loopback and we
traced it to an inversion of when we need to set MSG_MORE.  It should be
set when we have more bvec's to send, not when we are on the last bvec.
This patch made the test go from 20 iops to 78k iops.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Fixes: 429a787b ("nbd: fix use-after-free of rq/bio in the xmit path")
Signed-off-by: NJens Axboe <axboe@fb.com>

Additionally, don't assign directly to disk->queue, otherwise
blk_put_queue (called via put_disk) will choke (panic) on the errno
stored there.

Bug found by code inspection after Omar found a similar issue in
virtio_blk.  Compile-tested only.
Signed-off-by: NJeff Moyer <jmoyer@redhat.com>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Reviewed-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

... and don't mess with kmap() - just use BVEC_ITER for those parts.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Step 1 - don't reinintialize ->msg_iter on each iteration.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This was entirely automated, using the script by Al:

  PATT='^[[:blank:]]*#[[:blank:]]*include[[:blank:]]*<asm/uaccess.h>'
  sed -i -e "s!$PATT!#include <linux/uaccess.h>!" \
        $(git grep -l "$PATT"|grep -v ^include/linux/uaccess.h)

to do the replacement at the end of the merge window.
Requested-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

While doing stress tests we noticed that we'd get a lot of dmesg spam if
we suddenly disconnected the nbd device out of band.  Rate limit the
messages in the io path in order to deal with this.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

If an app exits before running NBD_DO_IT but after adding sockets we can
end up not being allowed to do a new nbd device.  Fix this by making
NBD_CLEAR_SOCK reset the setup_task.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We have this:

ERROR: "__aeabi_ldivmod" [drivers/block/nbd.ko] undefined!
ERROR: "__divdi3" [drivers/block/nbd.ko] undefined!
nbd.c:(.text+0x247c72): undefined reference to `__divdi3'

due to a recent commit, that did 64-bit division. Use the proper
divider function so that 32-bit compiles don't break.

Fixes: ef77b515 ("nbd: use loff_t for blocksize and nbd_set_size args")
Signed-off-by: NJens Axboe <axboe@fb.com>

If we have large devices (say like the 40t drive I was trying to test with) we
will end up overflowing the int arguments to nbd_set_size and not get the right
size for our device.  Fix this by using loff_t everywhere so I don't have to
think about this again.  Thanks,
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Multiple paths don't set it properly, ensure that we do.

Fixes: 9561a7ad ("nbd: add multi-connection support")
Signed-off-by: NJens Axboe <axboe@fb.com>

NBD can become contended on its single connection.  We have to serialize all
writes and we can only process one read response at a time.  Fix this by
allowing userspace to provide multiple connections to a single nbd device.  This
coupled with block-mq drastically increases performance in multi-process cases.
Thanks,
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

For writes, we can get a completion in while we're still iterating
the request and bio chain. If that happens, we're reading freed
memory and we can crash.

Break out after the last segment and avoid having the iterator
read freed memory.
Reviewed-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

'blk_mq_alloc_request()' returns an error pointer in case of error, not
NULL. So test it with IS_ERR.

Fixes: 	fd8383fd ("nbd: convert to blkmq")
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: NJens Axboe <axboe@fb.com>

Commit 0eadf37a ("nbd: allow block mq to deal with timeouts")
changed normal usage of nbd->sock_lock to use spin_lock/spin_unlock
rather than the *_irq variants, but it missed this unlock in an
error path.

Found by Coverity, CID 1373871.
Signed-off-by: NJohn W. Linville <linville@tuxdriver.com>
Cc: Josef Bacik <jbacik@fb.com>
Cc: Jens Axboe <axboe@fb.com>
Cc: Markus Pargmann <mpa@pengutronix.de>
Fixes: 0eadf37a ("nbd: allow block mq to deal with timeouts")
Signed-off-by: NJens Axboe <axboe@fb.com>

We take a mutex when sending commands and send stuff over the network, we need
to have queue_rq called asynchronously.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Fixes: fd8383fd ("nbd: convert to blkmq")
Signed-off-by: NJens Axboe <axboe@fb.com>

Instead of rolling our own timer, just utilize the blk mq req timeout and do the
disconnect if any of our commands timeout.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

In preparation for some future changes, change a few of the state bools over to
normal bits to set/clear properly.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We hit a warning when shutting down the nbd connection because we have irq's
disabled.  We don't really need to do the shutdown under the lock, just clear
the nbd->sock.  So do the shutdown outside of the irq.  This gets rid of the
warning.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This moves NBD over to using blkmq, which allows us to get rid of the NBD
wide queue lock and the async submit kthread.  We will start with 1 hw
queue for now, but I plan to add multiple tcp connection support in the
future and we'll fix how we set the hwqueue's.
Signed-off-by: NJosef Bacik <jbacik@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Quentin ran into this bug:

WARNING: CPU: 64 PID: 10085 at fs/sysfs/dir.c:31 sysfs_warn_dup+0x65/0x80
sysfs: cannot create duplicate filename '/devices/virtual/block/nbd3/pid'
Modules linked in: nbd
CPU: 64 PID: 10085 Comm: qemu-nbd Tainted: G      D         4.6.0+ #7
 0000000000000000 ffff8820330bba68 ffffffff814b8791 ffff8820330bbac8
 0000000000000000 ffff8820330bbab8 ffffffff810d04ab ffff8820330bbaa8
 0000001f00000296 0000000000017681 ffff8810380bf000 ffffffffa0001790
Call Trace:
 [<ffffffff814b8791>] dump_stack+0x4d/0x6c
 [<ffffffff810d04ab>] __warn+0xdb/0x100
 [<ffffffff810d0574>] warn_slowpath_fmt+0x44/0x50
 [<ffffffff81218c65>] sysfs_warn_dup+0x65/0x80
 [<ffffffff81218a02>] sysfs_add_file_mode_ns+0x172/0x180
 [<ffffffff81218a35>] sysfs_create_file_ns+0x25/0x30
 [<ffffffff81594a76>] device_create_file+0x36/0x90
 [<ffffffffa0000e8d>] __nbd_ioctl+0x32d/0x9b0 [nbd]
 [<ffffffff814cc8e8>] ? find_next_bit+0x18/0x20
 [<ffffffff810f7c29>] ? select_idle_sibling+0xe9/0x120
 [<ffffffff810f6cd7>] ? __enqueue_entity+0x67/0x70
 [<ffffffff810f9bf0>] ? enqueue_task_fair+0x630/0xe20
 [<ffffffff810efa76>] ? resched_curr+0x36/0x70
 [<ffffffff810f0078>] ? check_preempt_curr+0x78/0x90
 [<ffffffff810f00a2>] ? ttwu_do_wakeup+0x12/0x80
 [<ffffffff810f01b1>] ? ttwu_do_activate.constprop.86+0x61/0x70
 [<ffffffff810f0c15>] ? try_to_wake_up+0x185/0x2d0
 [<ffffffff810f0d6d>] ? default_wake_function+0xd/0x10
 [<ffffffff81105471>] ? autoremove_wake_function+0x11/0x40
 [<ffffffffa0001577>] nbd_ioctl+0x67/0x94 [nbd]
 [<ffffffff814ac0fd>] blkdev_ioctl+0x14d/0x940
 [<ffffffff811b0da2>] ? put_pipe_info+0x22/0x60
 [<ffffffff811d96cc>] block_ioctl+0x3c/0x40
 [<ffffffff811ba08d>] do_vfs_ioctl+0x8d/0x5e0
 [<ffffffff811aa329>] ? ____fput+0x9/0x10
 [<ffffffff810e9092>] ? task_work_run+0x72/0x90
 [<ffffffff811ba627>] SyS_ioctl+0x47/0x80
 [<ffffffff8185f5df>] entry_SYSCALL_64_fastpath+0x17/0x93
---[ end trace 7899b295e4f850c8 ]---

It seems fairly obvious that device_create_file() is not being protected
from being run concurrently on the same nbd.

Quentin found the following relevant commits:

1a2ad211 nbd: add locking to nbd_ioctl
90b8f282 [PATCH] end of methods switch: remove the old ones
d4430d62 [PATCH] beginning of methods conversion
08f85851 [PATCH] move block_device_operations to blkdev.h

It would seem that the race was introduced in the process of moving nbd
from BKL to unlocked ioctls.

By setting nbd->task_recv while the mutex is held, we can prevent other
processes from running concurrently (since nbd->task_recv is also checked
while the mutex is held).
Reported-and-tested-by: NQuentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Markus Pargmann <mpa@pengutronix.de>
Cc: Paul Clements <paul.clements@steeleye.com>
Cc: Pavel Machek <pavel@suse.cz>
Cc: Jens Axboe <axboe@fb.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NVegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: NJens Axboe <axboe@fb.com>