Add support for RBD_FEATURE_DATA_POOL feature.  rbd_dev->layout.pool_id
now stores the data pool id.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

Rather than initializing layout fields with some made up values in
__rbd_dev_create(), move the initialization into rbd_init_layout() and
call it after the header is actually populated.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

Returning u64 doesn't make sense: max header->obj_order is 25 and
ceph_file_layout::object_size is u32.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

As explained in the previous commit, rbd_obj_request machinery (and
rbd_osd_req_create() in particular) shouldn't be used for working with
metadata objects.

Switch to the recently added ceph_osdc_call().  It assumes single pages
for outbound and inbound buffers, but that's OK - none of the callers
need more than that.  These pages need to be allocated (messenger is in
dire need of proper iterator interface!), but we are swapping for
pages[] and pagelist allocations in the existing code.

Kill class_name argument - all rbd methods are under "rbd".
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

rbd_obj_request machinery is completely unnecessary here; all that's
being done is fetching a metadata object - no striping, cloning, etc.
More importantly, rbd_osd_req_create() grabs pool id from layout and
that is becoming a data pool id.

Kill offset argument - all metadata objects are small and read in full.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

No reason to delay it until image_id is known.  This will be required
by some rbd_obj_method_sync() callers, after rbd_obj_method_sync() is
changed to take oloc.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

Image format 1 is deprecated and format 2 doesn't have these.  Also,
__rbd_dev_create() takes care of zeroing (or otherwise initializing)
format 2 specific fields.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NJason Dillaman <dillaman@redhat.com>

drivers/block/rbd.c: In function ‘rbd_watch_cb’:
drivers/block/rbd.c:3690:5: error: ‘struct_v’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
drivers/block/rbd.c:3759:5: note: ‘struct_v’ was declared here
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

If the header object gets deleted (perhaps along with the entire pool),
there is no point in attempting to reregister the watch.  Treat this
the same as blacklisting: fail all pending and new I/Os requiring the
lock.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

-EBLACKLISTED from __rbd_register_watch() means that our ceph_client
got blacklisted - we won't be able to restore the watch and reacquire
the lock.  Wake up and fail all outstanding requests waiting for the
lock and arrange for all new requests that require the lock to fail
immediately.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Tested-by: NMike Christie <mchristi@redhat.com>

* A multiplication for the size determination of a memory allocation
  indicated that an array data structure should be processed.
  Thus use the corresponding function "kmalloc_array".

  This issue was detected by using the Coccinelle software.

* Delete the local variable "size" which became unnecessary with
  this refactoring.
Signed-off-by: NMarkus Elfring <elfring@users.sourceforge.net>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Pull setting an error and marking a request done code into a new
helper.  obj_request_img_data_test() check isn't strictly needed right
now, but makes it applicable to !img_data requests and a bit safer.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Move the check into rbd_obj_request_destroy() to avoid use-after-free
on errors in rbd_img_request_fill(..., OBJ_REQUEST_PAGES, ...), where
pages, owned by the caller, gets freed in rbd_img_request_fill().
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

Accessing obj_request->img_request union field is only valid for object
requests associated with an image (i.e. if obj_request_img_data_test()
returns true).  rbd_osd_req_format_read() used to do more, but now it
just sets osd_req->snap_id.  Standalone and stat object requests always
go to the HEAD revision and are fine with CEPH_NOSNAP set by libceph,
so get around the invalid union field use by simply not calling
rbd_osd_req_format_read() in those places.
Reported-by: NDavid Disseldorp <ddiss@suse.de>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

- don't put obj_request before rbd_obj_request_get() if
  rbd_obj_request_create() fails
- don't leak pages if rbd_obj_request_create() fails
- don't leak stat_request if rbd_osd_req_create() fails
Reported-by: NDavid Disseldorp <ddiss@suse.de>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

- fix parent_length == img_request->xferred assert to not fire on
  copyup read failures
- don't leak pages if copyup read fails or we can't allocate a new osd
  request
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

Commit 0f2d5be7 ("rbd: use reference counts for image requests")
added rbd_img_request_get(), which rbd_img_request_fill() calls for
each obj_request added to img_request.  It was an urgent band-aid for
the uglyness that is rbd_img_obj_callback() and none of the error paths
were updated.

Given that this img_request reference is meant to represent an
obj_request that hasn't passed through rbd_img_obj_callback() yet,
proper cleanup in appropriate destructors is a challenge.  However,
noting that if we don't get a chance to call rbd_obj_request_complete(),
there is not going to be a call to rbd_img_obj_callback(), we can move
rbd_img_request_get() into rbd_obj_request_submit() and fixup the two
places that call rbd_obj_request_complete() directly and not through
rbd_obj_request_submit() to temporarily bump img_request, so that
rbd_img_obj_callback() can put as usual.

This takes care of img_request leaks on errors on the submit side.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>

If stat request fails with something other than -ENOENT (which just
means that we need to copyup), the original object request is never
marked as done and therefore never completed.  Fix this by moving the
mark done + complete snippet from rbd_img_obj_parent_read_full() into
rbd_img_obj_exists_callback().  The former remains covered, as the
latter is its only caller (through rbd_img_obj_request_submit()).
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

Assert once in rbd_img_obj_request_submit().
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

- osdc parameter is useless
- starting with commit 5aea3dcd ("libceph: a major OSD client
  update"), ceph_osdc_start_request() always returns success
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>
Reviewed-by: NDavid Disseldorp <ddiss@suse.de>

Add a per-device option to acquire exclusive lock on reads (in addition
to writes and discards).  The use case is iSCSI, where it will be used
to prevent execution of stale writes after the implicit failover.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Tested-by: NMike Christie <mchristi@redhat.com>

All drivers use the default, so provide an inline version of it.  If we
ever need other queue mapping we can add an optional method back,
although supporting will also require major changes to the queue setup
code.

This provides better code generation, and better debugability as well.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This adds a force close option, so we can force the unmapping
of a rbd device that is open. If a path/device is blacklisted, apps
like multipathd can map a new device and then unmap the old one.
The unmapping cleanup would then be handled by the generic hotunplug
code paths in multipahd like is done for iSCSI, FC/FCOE, SAS, etc.
Signed-off-by: NMike Christie <mchristi@redhat.com>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Export the info used to setup the rbd image, so it can be used to remap
the image.
Signed-off-by: NMike Christie <mchristi@redhat.com>
[idryomov@gmail.com: do_rbd_add() EH]
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Export snap id in sysfs, so tools like multipathd can use it in a uuid.
Signed-off-by: NMike Christie <mchristi@redhat.com>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Export the cluster fsid, so tools like udev and multipath-tools can use
it for part of the uuid.
Signed-off-by: NMike Christie <mchristi@redhat.com>
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Export client addr/nonce, so userspace can check if a image is being
blacklisted.
Signed-off-by: NMike Christie <mchristi@redhat.com>
[idryomov@gmail.com: ceph_client_addr(), endianess fix]
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

With exclusive-lock added and more to come, print features into dmesg.
Change capacity to decimal while at it.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NMike Christie <mchristi@redhat.com>

Add basic support for RBD_FEATURE_EXCLUSIVE_LOCK feature.  Maintenance
operations (resize, snapshot create, etc) are offloaded to librbd via
returning -EOPNOTSUPP - librbd should request the lock and execute the
operation.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NMike Christie <mchristi@redhat.com>
Tested-by: NMike Christie <mchristi@redhat.com>

Revamp watch code to support retrying watch re-registration:

- add rbd_dev->watch_state for more robust errcb handling
- store watch cookie separately to avoid dereferencing watch_handle
  which is set to NULL on unwatch
- move re-register code into a delayed work and retry re-registration
  every second, unless the client is blacklisted
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NMike Christie <mchristi@redhat.com>
Tested-by: NMike Christie <mchristi@redhat.com>

This is going to be used for re-registering watch requests and
exclusive-lock tasks: acquire/request lock, notify-acquired, release
lock, notify-released.  Some refactoring in the map/unmap paths was
necessary to give this workqueue a meaningful name: "rbdX-tasks".
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NMike Christie <mchristi@redhat.com>

It's gid / global_id in other places.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NMike Christie <mchristi@redhat.com>
Reviewed-by: NAlex Elder <elder@linaro.org>

ceph_file_layout::pool_id is now s64.  rbd_add_get_pool_id() and
ceph_pg_poolid_by_name() both return an int, so it's bogus anyway.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>
Reviewed-by: NAlex Elder <elder@linaro.org>

Purely cosmetic at this point, as rbd doesn't use RADOS namespaces and
hence rbd_dev->header_oloc->pool_ns is always NULL.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Add pool namesapce pointer to struct ceph_file_layout and struct
ceph_object_locator. Pool namespace is used by when mapping object
to PG, it's also used when composing OSD request.

The namespace pointer in struct ceph_file_layout is RCU protected.
So libceph can read namespace without taking lock.
Signed-off-by: NYan, Zheng <zyan@redhat.com>
[idryomov@gmail.com: ceph_oloc_destroy(), misc minor changes]
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>

Define new ceph_file_layout structure and rename old ceph_file_layout
to ceph_file_layout_legacy. This is preparation for adding namespace
to ceph_file_layout structure.
Signed-off-by: NYan, Zheng <zyan@redhat.com>

The req operation REQ_OP is separated from the rq_flag_bits
definition. This converts the block layer drivers to
use req_op to get the op from the request struct.
Signed-off-by: NMike Christie <mchristi@redhat.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

... with a wrapper around maybe_request_map() - no need for two
osdmap-specific functions.
Signed-off-by: NIlya Dryomov <idryomov@gmail.com>