Similar to userspace, don't bail with "parse_ips bad ip ..." if the
specified port is port 0, instead use port CEPH_MON_PORT (6789, the
default monitor port).
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Add CRUSH_V2 feature (new indep mode and SET_* steps) to a set of
features supported by default.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit 8b38f10bc2ee3643a33ea5f9545ad5c00e4ac5b4.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit ea3a0bb8b773360d73b8b77fa32115ef091c9857.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

This allows all of the tunables to be overridden by a specific rule.

Reflects ceph.git commits d129e09e57fbc61cfd4f492e3ee77d0750c9d292,
                          0497db49e5973b50df26251ed0e3f4ac7578e66e.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

The legacy behavior is to make the normal number of tries for the
recursive chooseleaf call.  The descend_once tunable changed this to
making a single try and bail if we get a reject (note that it is
impossible to collide in the recursive case).

The new set_chooseleaf_tries lets you select the number of recursive
chooseleaf attempts for indep mode, or default to 1.  Use the same
behavior for firstn, except default to total_tries when the legacy
tunables are set (for compatibility).  This makes the rule step
override the (new) default of 1 recursive attempt, keeping behavior
consistent with indep mode.

Reflects ceph.git commit 685c6950ef3df325ef04ce7c986e36ca2514c5f1.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

This aligns the internal identifier names with the user-visible names in
the decompiled crush map language.

Reflects ceph.git commit caa0e22e15e4226c3671318ba1f61314bf6da2a6.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Since we can specify the recursive retries in a rule, we may as well also
specify the non-recursive tries too for completeness.

Reflects ceph.git commit d1b97462cffccc871914859eaee562f2786abfd1.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Parameterize the attempts for the _firstn choose method, and apply the
rule-specified tries count to firstn mode as well.  Note that we have
slightly different behavior here than with indep:

 If the firstn value is not specified for firstn, we pass through the
 normal attempt count.  This maintains compatibility with legacy behavior.
 Note that this is usually *not* actually N^2 work, though, because of the
 descend_once tunable.  However, descend_once is unfortunately *not* the
 same thing as 1 chooseleaf try because it is only checked on a reject but
 not on a collision.  Sigh.

 In contrast, for indep, if tries is not specified we default to 1
 recursive attempt, because that is simply more sane, and we have the
 option to do so.  The descend_once tunable has no effect for indep.

Reflects ceph.git commit 64aeded50d80942d66a5ec7b604ff2fcbf5d7b63.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Explicitly control the number of sample attempts, and allow the number of
tries in the recursive call to be explicitly controlled via the rule. This
is important because the amount of time we want to spend looking for a
solution may be rule dependent (e.g., higher for the wide indep pool than
the rep pools).

(We should do the same for the other tunables, by the way!)

Reflects ceph.git commit c43c893be872f709c787bc57f46c0e97876ff681.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Pass down the parent's 'r' value so that we will sample different values in
the recursive call when the parent tries multiple times.  This avoids doing
useless work (calling multiple times and trying the same values).

Reflects ceph.git commit 2731d3030d7a3e80922b7f1b7756f9a4a124bac5.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Pass numrep (the width of the result) separately from the number of results
we want *this* iteration.  This makes things less awkward when we do a
recursive call (for chooseleaf) and want only one item.

Reflects ceph.git commit 1b567ee08972f268c11b43fc881e57b5984dd08b.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Now that indep is handled by crush_choose_indep, rename crush_choose to
crush_choose_firstn and remove all the conditionals.  This ends up
stripping out *lots* of code.

Note that it *also* makes it obvious that the shenanigans we were playing
with r' for uniform buckets were broken for firstn mode.  This appears to
have happened waaaay back in commit dae8bec9 (or earlier)... 2007.

Reflects ceph.git commit 94350996cb2035850bcbece6a77a9b0394177ec9.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit 4551fee9ad89d0427ed865d766d0d44004d3e3e1.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit 86e978036a4ecbac4c875e7c00f6c5bbe37282d3.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

For firstn mode, if we fail to make a valid placement choice, we just
continue and return a short result to the caller.  For indep mode, however,
we need to make the position stable, and return an undefined value on
failed placements to avoid shifting later results to the left.

Reflects ceph.git commit b1d4dd4eb044875874a1d01c01c7d766db5d0a80.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

This is only present to size the temporary scratch arrays that we put on
the stack.  Let the caller allocate them as they wish and remove the
limitation.

Reflects ceph.git commit 1cfe140bf2dab99517589a82a916f4c75b9492d1.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit 3cef755428761f2481b1dd0e0fbd0464ac483fc5.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit e7d47827f0333c96ad43d257607fb92ed4176550.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Reflects ceph.git commit 43a01c9973c4b83f2eaa98be87429941a227ddde.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Pass the size of the weight vector into crush_do_rule() to ensure that we
don't access values past the end.  This can happen if the caller misbehaves
and passes a weight vector that is smaller than max_devices.

Currently the monitor tries to prevent that from happening, but this will
gracefully tolerate previous bad osdmaps that got into this state.  It's
also a bit more defensive.

Reflects ceph.git commit 5922e2c2b8335b5e46c9504349c3a55b7434c01a.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

This updates ceph_features.h so that it has all feature bits defined in
ceph.git.  In the interim since the last update, ceph.git crossed the
"32 feature bits" point, and, the addition of the 33rd bit wasn't
handled correctly.  The work-around is squashed into this commit and
reflects ceph.git commit 053659d05e0349053ef703b414f44965f368b9f0.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

In preparation for ceph_features.h update, change all features fields
from unsigned int/u32 to u64.  (ceph.git has ~40 feature bits at this
point.)
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Tear down watch request if rbd_dev_device_setup() fails.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Rename rbd_dev_header_watch_sync() to __rbd_dev_header_watch_sync() and
introduce two helpers: rbd_dev_header_{,un}watch_sync() to make it more
clear what is going on.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

I no longer have direct access to my Inktank e-mail.  I still pay
attention to rbd, so update its entry in MAINTAINERS accordingly.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Sage Weil <sage@inktank.com>

If single-major device number allocation scheme is turned on, instead
of reserving 256 minors per device, which imposes a limit of 4096
images mapped at once, reserve 16 minors per device and enable extended
devt feature.  This results in a theoretical limit of 65536 images
mapped at once, and still allows to have more than 15 partititions:
partitions starting with 16th are mapped under major 259 (Block
Extended Major):

$ rbd showmapped
id pool image snap device
0  rbd  b5    -    /dev/rbd0    # no partitions
1  rbd  b2    -    /dev/rbd1    # 40 partitions
2  rbd  b3    -    /dev/rbd2    #  2 partitions

$ cat /proc/partitions
 251        0       1024 rbd0
 251       16       1024 rbd1
 251       17          0 rbd1p1
 251       18          0 rbd1p2
 ...
 251       30          0 rbd1p14
 251       31          0 rbd1p15
 259        0          0 rbd1p16
 259        1          0 rbd1p17
 ...
 259       23          0 rbd1p39
 259       24          0 rbd1p40
 251       32       1024 rbd2
 251       33          0 rbd2p1
 251       34          0 rbd2p2

(major 251 was assigned dynamically at module load time)
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Currently, if one new page allocated into fscache in readpage(), however,
with no data read into due to error encountered during reading from OSDs,
the slot in fscache is not uncached. This patch fixes this.
Signed-off-by: Li Wang <liwang@ubuntukylin.com>
Reviewed-by: Milosz Tanski <milosz@adfin.com>

Signed-off-by: Li Wang <liwang@ubuntukylin.com>
Reviewed-by: Milosz Tanski <milosz@adfin.com>

Signed-off-by: Guangliang Zhao <lucienchao@gmail.com>
Reviewed-by: Li Wang <li.wang@ubuntykylin.com>
Reviewed-by: Zheng Yan <zheng.z.yan@intel.com>

Adds cap check to the page fault handler. The check prevents page
fault handler from adding new page to the page cache while Fcb caps
are being revoked. This solves Fc revoking hang in multiple clients
mmap IO workload.
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Reviewed-by: Sage Weil <sage@inktank.com>

Currently each rbd device is allocated its own major number, which
leads to a hard limit of 230-250 images mapped at once.  This commit
adds support for a new single-major device number allocation scheme,
which is hidden behind a new single_major boolean module parameter and
is disabled by default for backwards compatibility reasons.  (Old
userspace cannot correctly unmap images mapped under single-major
scheme and would essentially just unmap a random image, if that.)

$ rbd showmapped
id pool image snap device
0  rbd  b100  -    /dev/rbd0
1  rbd  b101  -    /dev/rbd1
2  rbd  b102  -    /dev/rbd2
3  rbd  b103  -    /dev/rbd3

Old scheme (modprobe rbd):

$ ls -l /dev/rbd*
brw-rw---- 1 root disk 253, 0 Dec 10 12:24 /dev/rbd0
brw-rw---- 1 root disk 252, 0 Dec 10 12:28 /dev/rbd1
brw-rw---- 1 root disk 252, 1 Dec 10 12:28 /dev/rbd1p1
brw-rw---- 1 root disk 252, 2 Dec 10 12:28 /dev/rbd1p2
brw-rw---- 1 root disk 252, 3 Dec 10 12:28 /dev/rbd1p3
brw-rw---- 1 root disk 251, 0 Dec 10 12:28 /dev/rbd2
brw-rw---- 1 root disk 251, 1 Dec 10 12:28 /dev/rbd2p1
brw-rw---- 1 root disk 250, 0 Dec 10 12:24 /dev/rbd3

New scheme (modprobe rbd single_major=Y):

$ ls -l /dev/rbd*
brw-rw---- 1 root disk 253,   0 Dec 10 12:30 /dev/rbd0
brw-rw---- 1 root disk 253, 256 Dec 10 12:30 /dev/rbd1
brw-rw---- 1 root disk 253, 257 Dec 10 12:30 /dev/rbd1p1
brw-rw---- 1 root disk 253, 258 Dec 10 12:30 /dev/rbd1p2
brw-rw---- 1 root disk 253, 259 Dec 10 12:30 /dev/rbd1p3
brw-rw---- 1 root disk 253, 512 Dec 10 12:30 /dev/rbd2
brw-rw---- 1 root disk 253, 513 Dec 10 12:30 /dev/rbd2p1
brw-rw---- 1 root disk 253, 768 Dec 10 12:30 /dev/rbd3

(major 253 was assigned dynamically at module load time)

The new limit is 4096 images mapped at once, and it comes from the fact
that, as before, 256 minor numbers are reserved for each mapping.
(A follow-up commit changes the number of minors reserved and the way
we deal with partitions over that number.)

If single_major is set to true, two new sysfs interfaces show up:
/sys/bus/rbd/{add,remove}_single_major.  These are to be used instead
of /sys/bus/rbd/{add,remove}, which are disabled for backwards
compatibility reasons outlined above.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

In preparation for single-major device number allocation scheme, wire
up attribute_group::is_visible() callback for rbd bus.  This allows us
to make the new single-major attributes conditional.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Introduce /sys/bus/rbd/devices/<id>/minor sysfs attribute for exporting
rbd whole disk minor numbers.  This is a step towards single-major
device number allocation scheme, but also a good thing on its own.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Currently rbd ids are allocated using an atomic variable that keeps
track of the highest id currently in use and each new id is simply one
more than the value of that variable.  That's nice and cheap, but it
does mean that rbd ids are allowed to grow boundlessly, and, more
importantly, it's completely unpredictable.  So, in preparation for
single-major device number allocation scheme, which is going to
establish and rely on a constant mapping between rbd ids and device
numbers, switch to ida for rbd id assignments.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Refactor rbd_init() a bit to make it more clear what's going on.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

Tweak "loaded" message, so that it looks like

[   30.184235] rbd: loaded

instead of

[   38.056564] rbd: loaded rbd (rados block device)

Also move (and slightly tweak) MODULE_DESCRIPTION so that all authors
are next to each other in modinfo output.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

rbd_device::dev_id is an int, format it as such.
Signed-off-by: Ilya Dryomov <ilya.dryomov@inktank.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>

With the current full handling, there is a race between osds and
clients getting the first map marked full. If the osd wins, it will
return -ENOSPC to any writes, but the client may already have writes
in flight. This results in the client getting the error and
propagating it up the stack. For rbd, the block layer turns this into
EIO, which can cause corruption in filesystems above it.

To avoid this race, osds are being changed to drop writes that came
from clients with an osdmap older than the last osdmap marked full.
In order for this to work, clients must resend all writes after they
encounter a full -> not full transition in the osdmap. osds will wait
for an updated map instead of processing a request from a client with
a newer map, so resent writes will not be dropped by the osd unless
there is another not full -> full transition.

This approach requires both osds and clients to be fixed to avoid the
race. Old clients talking to osds with this fix may hang instead of
returning EIO and potentially corrupting an fs. New clients talking to
old osds have the same behavior as before if they encounter this race.

Fixes: http://tracker.ceph.com/issues/6938Reviewed-by: Sage Weil <sage@inktank.com>
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>

The PAUSEWR and PAUSERD flags are meant to stop the cluster from
processing writes and reads, respectively. The FULL flag is set when
the cluster determines that it is out of space, and will no longer
process writes.  PAUSEWR and PAUSERD are purely client-side settings
already implemented in userspace clients. The osd does nothing special
with these flags.

When the FULL flag is set, however, the osd responds to all writes
with -ENOSPC. For cephfs, this makes sense, but for rbd the block
layer translates this into EIO.  If a cluster goes from full to
non-full quickly, a filesystem on top of rbd will not behave well,
since some writes succeed while others get EIO.

Fix this by blocking any writes when the FULL flag is set in the osd
client. This is the same strategy used by userspace, so apply it by
default.  A follow-on patch makes this configurable.

__map_request() is called to re-target osd requests in case the
available osds changed.  Add a paused field to a ceph_osd_request, and
set it whenever an appropriate osd map flag is set.  Avoid queueing
paused requests in __map_request(), but force them to be resent if
they become unpaused.

Also subscribe to the next osd map from the monitor if any of these
flags are set, so paused requests can be unblocked as soon as
possible.

Fixes: http://tracker.ceph.com/issues/6079Reviewed-by: Sage Weil <sage@inktank.com>
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>