Fix casing of Content-Type header

It turns out, despite the HTTP spec declaring that header field names
should be case-insensitive, some clients treat them wrongly, and
consider "Content-type" to not match "Content-Type".

CyberDuck was one of those clients, now fixed upstream in
https://trac.cyberduck.io/ticket/8999

To reduce future occurances of this bug, fix the casing of the
Content-Type header, to strictly comply with the HTTP specification (be
strict about what you send, and generous about what you receive).

Fixes: #12939
Backport: hammer, firefly
Signed-off-by: Robin H. Johnson <robin.johnson@dreamhost.com>
(cherry picked from commit 1b9fbffdc24160251b96cec820d62fb2a12b6eab)

rgw: we should not overide Swift sent content type

Fixes: #12363
backport: hammer

Signed-off-by: Orit Wasserman <owasserm@redhat.com>
(cherry picked from commit 423cf136f15df3099c9266f55932542d303c2713)

rgw: send Content-Length in response for GET on Swift account.

Fixes: #12158
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
(cherry picked from commit 5f47b11791cd46ce2771046e4888575869cc9685)

Conflicts:
src/rgw/rgw_rest_swift.cc
            omit dump_account_metadata() call

rgw: enforce Content-Type in Swift responses.

Swift sends Content-Type HTTP header even if the response
doesn't contain body. We have this behaviour implemented
until applying some changes in end_header() function.
Unfortunately, lack of Content-Type causes early exits in
many Tempest's tests for Swift API verification.

Fixes: #12157
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
(cherry picked from commit 106aeba206736d4080326f9bc191876bed63370b)

rgw: force content_type for swift bucket stats request

Fixes: 12095
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
(cherry picked from commit da259e4da1329e1e448d75c48c6ee3d1318e19b2)

rgw: force content-type header for swift account responses without body

This regression was introduced in commit 4cfeca03ade21861ab70ca759f94eb244c16cb39

Fixes: #11438
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
(cherry picked from commit 009664e560ee37ef6dc170a0155d6c749e151b06)

rgw: shouldn't return content-type: application/xml if content length is 0

Fixed: #11091

Signed-off-by: Orit Wasserman <owasserm@redhat.com>
(cherry picked from commit 5cc92bb27cfbaee4af24dc312c9b98331f899794)

Merge pull request #5993 from SUSE/wip-13169-firefly

update docs to point to download.ceph.com

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #6038 from liewegas/wip-13088

common/HeartBeat: include types

Reviewed-by: Kefu Chai <kchai@redhat.com>

common/RWLock: include assert

Fixes atomic_t build issue without libatomic-ops.

Fixes: #13088
Signed-off-by: Sage Weil <sage@redhat.com>

.gitmodules: fix ceph-object-corpus source url

Fixes: #13196
Signed-off-by: Sage Weil <sage@redhat.com>

Merge pull request #6000 from SUSE/wip-12097-firefly

kernel_untar_build fails on EL7

Merge pull request #5234 from SUSE/wip-12244-firefly

Civetweb RGW appears to report full size of object as downloaded when only partially downloaded

Merge pull request #4535 from xinxinsh/wip-11416-firefly

rgw: use correct objv_tracker for bucket instance

qa: update to newer Linux tarball

This should make newer gcc releases happier in their default configuration.
kernel.org is now distributing tarballs as .xz files so we change to that
as well when decompressing (it is supported by Ubuntu Precise so we should
be all good).

Fixes: #11758
Signed-off-by: Greg Farnum <gfarnum@redhat.com>
(cherry picked from commit 1ea3f47ab806d48ca7b045c2731d344eae3900e1)

Conflicts:
qa/workunits/kernel_untar_build.sh
            trivial resolution

doc: correct links to download.ceph.com

Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit 387d7800359154950431d0984c756f43f21dd9b4)

Conflicts:
doc/install/get-packages.rst
            preferred the new version in all cases

Merge pull request #5200 from SUSE/wip-12289-firefly

mon: scrub error (osdmap encoding mismatch?) upgrading from 0.80 to ~0.80.2

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5388 from SUSE/wip-12490-firefly

buffer: critical bufferlist::zero bug

Reviewed-by: Gregory Farnum <gfarnum@redhat.com>

Merge pull request #5408 from SUSE/wip-12492-firefly

the output is wrong when runing ceph osd reweight

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5404 from SUSE/wip-12395-firefly

register_new_pgs() should check ruleno instead of its index

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5199 from SUSE/wip-11980-firefly

MDSMonitor: handle MDSBeacon messages properly

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5410 from SUSE/wip-12497-firefly

get pools health'info have error

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5409 from SUSE/wip-12495-firefly

pgmonitor: wrong at/near target max“ reporting

Reviewed-by: Kefu Chai <kchai@redhat.com>

Merge pull request #5358 from ceph/wip-11470.firefly

mon: PaxosService: call post_refresh() instead of post_paxos_update()

Reviewed-by: Joao Eduardo Luis <joao@suse.de>

Merge pull request #5403 from SUSE/wip-12393-firefly

Memory leak in Mutex.cc, pthread_mutexattr_init without pthread_mutexattr_destroy

Reviewed-by: Kefu Chai <kchai@redhat.com>
Reviewed-by: Nathan Cutler <ncutler@suse.com>

Merge pull request #5225 from SUSE/wip-12266-firefly

ceph.spec.in: 50-rbd.rules conditional is wrong

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5217 from SUSE/wip-12268-firefly

ceph-common needs python-argparse on older distros, but doesn't require it

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5224 from SUSE/wip-12304-firefly

ceph.spec.in running fdupes unnecessarily

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #5394 from SUSE/wip-12447-firefly

ceph.spec.in: useless %py_requires breaks SLE11-SP3 build

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #5043 from SUSE/wip-12007-firefly

Objecter: pg listing can deadlock when throttling is in use

Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4769 from SUSE/wip-11741-firefly

crush: take crashes due to invalid arg

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4788 from SUSE/wip-11801-firefly

Make it more difficult to delete pools in firefly

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #5235 from SUSE/wip-12310-firefly

read on chunk-aligned xattr not handled

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5389 from SUSE/wip-12391-firefly

EC pools are not allowed as cache pools, disallow in the mon

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5406 from ceph/wip-12465-firefly

log: take mutex while opening fd

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #4854 from ceph/wip-11769-firefly

librbd: aio calls may block

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #5233 from SUSE/wip-12074-firefly

RGW logfile does not get chowned properly

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5390 from SUSE/wip-12503-firefly

rgw/logrotate.conf calls service with wrong init script name

Reviewed-by: Ken Dreyer <kdreyer@redhat.com>

Merge pull request #4765 from SUSE/wip-11721-firefly

rgw: Keystone PKI token expiration is not enforced

Update OSDMonitor.cc

OSDMonitor::get_pools_health(), s/objects/bytes/

Fixes: #12402
Signed-off-by: ren.huanwen@zte.com.cn
(cherry picked from commit 7fc13c9d6b9a4962d7640240416105d8f558d600)

  mon/PGMonitor: avoid uint64_t overflow when checking pool 'target/max' status.
Fixes: #12401
Signed-off-by: huangjun <hjwsm1989@gmail.com>
(cherry picked from commit f84e6b8157e78fd05cd1c8ac957c92df17e6ba06)

mon: OSDMonitor: fix hex output on 'osd reweight'

Fixes: #12251
Signed-off-by: Joao Eduardo Luis <joao@suse.de>
(cherry picked from commit 787fa80c2746fde44ac0583ff7995ec5be9a672d)

mon/PGMonitor: bug fix pg monitor get crush rule

when some rules have been deleted before, the index in array of crush->rules
is not always equals to crush_ruleset of pool.

Fixes: #12210
Reported-by: Ning Yao <zay11022@gmail.com>
Signed-off-by: Xinze Chi <xmdxcxz@gmail.com>
(cherry picked from commit 498793393c81c0a8e37911237969fba495a3a183)

Mutex: fix leak of pthread_mutexattr

Need pthread_mutexattr_destroy after pthread_mutexattr_init

Fixes: #111762
Signed-off-by: Ketor Meng <d.ketor@gmail.com>
(cherry picked from commit 2b23327b3aa8d96341d501a5555195ca1bc0de8f)

ceph.spec.in: drop SUSE-specific %py_requires macro

%py_requires expands to

BuildRequires: /usr/bin/python
PreReq: python = 2.7

The BuildRequires: is already provided, and the PreReq is wrong because
e.g. SLE11-SP3 (a platform we are trying to support) has Python 2.6.

http://tracker.ceph.com/issues/12351 Fixes: #12351

Signed-off-by: Nathan Cutler <ncutler@suse.com>
(cherry picked from commit cbf6c7b46b42e163950bf7956e5b7565fca961c9)

Conflicts:
ceph.spec.in
            in master, the %py_requires is in the python-rados package,
            while in firefly it is in the python-ceph package

rgw/logrotate.conf: Rename service name

The service name for ceph rados gateway was changed to "ceph-radosgw",
the previous version of service name "radosgw" would cause a failed reload,
and finally make it impossible to write any log data to the log file.

Signed-off-by: wuxingyi <wuxingyi2015@outlook.com>
(cherry picked from commit 9df3f798179481fe8ae6ae873dcb793de7d8f367)

OSDMonitor: disallow ec pools as tiers

Fixes: 11650
Signed-off-by: Samuel Just <sjust@redhat.com>
(cherry picked from commit 11b7801bb57cb25cd2d26d58722d49691747725b)

Conflicts:
qa/workunits/cephtool/test.sh
           no "# make sure we can't clobber snapshot state" tests in firefly
src/mon/OSDMonitor.cc
           no tp->removed_snaps.empty() in firefly

UnittestBuffer: Add bufferlist zero test case

Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
(cherry picked from commit 577acf665948e531e50d36f5780ea19351a5bf65)

buffer: Fix bufferlist::zero bug with special case

Fixes: #12252
Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
(cherry picked from commit 43f583d160ccaf879eaf0f3020e77860cf8d1df0)

mon: PaxosService: call post_refresh() instead of post_paxos_update()

Whenever the monitor finishes committing a proposal, we call
Monitor::refresh_from_paxos() to nudge the services to refresh.  Once
all services have refreshed, we would then call each services
post_paxos_update().

However, due to an unfortunate, non-critical bug, some services (mainly
the LogMonitor) could have messages pending in their
'waiting_for_finished_proposal' callback queue [1], and we need to nudge
those callbacks.

This patch adds a new step during the refresh phase: instead of calling
directly the service's post_paxos_update(), we introduce a
PaxosService::post_refresh() which will call the services
post_paxos_update() function first and then nudge those callbacks when
appropriate.

[1] - Given the monitor will send MLog messages to itself, and given the
service is not readable before its initial state is proposed and
committed, some of the initial MLog's would be stuck waiting for the
proposal to finish.  However, by design, we only nudge those message's
callbacks when an election finishes or, if the leader, when the proposal
finishes.  On peons, however, we would only nudge those callbacks if an
election happened to be triggered, hence the need for an alternate path
to retry any message waiting for the initial proposal to finish.

Fixes: #11470
Signed-off-by: Joao Eduardo Luis <joao@suse.de>
(cherry picked from commit 1551ebb63238073d2fd30201e6b656a8988e958c)

Log::reopen_log_file: take m_flush_mutex

Otherwise, _flush() might continue to write to m_fd after it's closed.
This might cause log data to go to a data object if the filestore then
reuses the fd during that time.

Fixes: #12465
Backport: firefly, hammer
Signed-off-by: Samuel Just <sjust@redhat.com>
(cherry picked from commit 8778ab3a1ced7fab07662248af0c773df759653d)

  - modified to drop m_flush_mutex_holder, which isn't present in firefly

Merge pull request #4583 from SUSE/wip-11542-firefly

rgw: Change variable length array of std::strings (not legal in C++) to std::vector<std::string>

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #4584 from SUSE/wip-8911-firefly

rgw: doesn't return 'x-timestamp' in header which is used by 'View Details' of OpenStack

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #4642 from SUSE/wip-11323-firefly

rgw: quota not respected in POST object

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

Merge pull request #4641 from SUSE/wip-11149-firefly

rgw: delete bucket does not remove .bucket.meta file

Reviewed-by: Yahuda Sadeh <yehuda@redhat.com>

Merge pull request #5170 from SUSE/wip-12039-firefly

rgw: Swift API: response for PUT on /container does not contain the mandatory Content-Length header when FCGI is used

Reviewed-by: Yahuda Sadeh <yehuda@redhat.com>

Merge pull request #4762 from SUSE/wip-11754-firefly

rgw: Object copy bug

Reviewed-by: Yahuda Sadeh <yehuda@redhat.com>

Merge pull request #5171 from SUSE/wip-12240-firefly

librbd: internal.cc: 1967: FAILED assert(watchers.size() == 1)

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

Merge pull request #4639 from SUSE/wip-7385-firefly

librbd: Objectcacher setting max object counts too low

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

tests: verify librbd blocking aio code path

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 4cf41486e9c9e1efcb863960a8f3e0326ffca7e5)

librbd: new rbd_non_blocking_aio config option

Setting this option to false reverts librbd to legacy behavior
where AIO operations could potentially block.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 769cad12716b85d87eacc1069dd9f5c21cad3915)

Conflicts:
src/common/config_opts.h: trivial resolution
src/librbd/librbd.cc: trivial resolution

PendingReleaseNotes: document changes to librbd's aio_read methods

Signed-off-by: Jason Dillaman <dillaman@redhat.com>

tests: update librbd AIO tests to remove result code

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
Conflicts:
src/test/librbd/test_librbd.cc: trivial resolution

librbd: AioRequest::send no longer returns a result

The librados calls used by AioRequest::send should always return
zero unless there is a bug.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit c77bce3311ab62892eb8c1d883263ba7ed663b20)

Conflicts:
src/librbd/AioRequest.cc: trivial resolution
src/librbd/AioRequest.h: trivial resolution
src/librbd/internal.cc: trivial resolution

librbd: internal AIO methods no longer return result

All failures should be returned via the AioCompletion.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 9ab42d613128ab08c688ddbea93df4c95068b9cd)

Conflicts:
src/librbd/AioRequest.cc: trivial resolution
src/librbd/internal.cc: trivial resolution
src/librbd/internal.h: trivial resolution

Throttle: added pending_error method to SimpleThrottle

Allow the client of SimpleThrottle to detect an async error
so that it can exit early.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit b88b88c5df91325fb713c2031a56bffe421268e0)

librbd: add new fail method to AioCompletion

Helper method to handle passing fatal errors generated within
librbd (not from the OSDs) back to the client.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 6d1d0c867855a96bee4c13a0c0a39a0e002ccd12)

Conflicts:
src/librbd/AioCompletion.cc: trivial resolution
src/librbd/AioCompletion.h: trivial resolution

librbd: avoid blocking AIO API methods

Enqueue all AIO API methods within the new librbd thread pool to
reduce the possibility of any blocking operations. To maintain
backwards compatibility with the legacy return codes of the API's
AIO methods, it's still possible to block attempting to acquire
the snap_lock.

Fixes: #11056
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 3a7b5e30efdb21aa1a0aeb68a5d02a1ac2a5faf3)

Conflicts:
src/librbd/librbd.cc: trivial resolution

librbd: add task pool / work queue for AIO requests

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit afb896d91f886b647baf38f7ec94cc3739f6d2a9)

Conflicts:
src/common/config_opts.h: trivial resolution
src/librbd/ImageCtx.cc: trivial resolution
src/librbd/ImageCtx.h: trivial resolution
src/librbd/internal.cc: trivial resolution

WorkQueue: added virtual destructor

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit b3f5a75332c058816dc39b71e9d2b36e752159f4)

WorkQueue: add new ContextWQ work queue

The queue holds a collection of Context pointers that will
be completed by the thread pool.

Signed-off-by: Jason Dillaman <dillaman@redhat.com>
(cherry picked from commit 24a33e977f7b71962adeeb48f75d488a76e70fa9)

Conflicts:
src/common/WorkQueue.h: trivial resolution

common/ceph_context: don't import std namespace

This was broken by 7fed5dee4f96a83d1d6914f6fc0895bba2d15b99

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 9029813b2784fe71879b1b7684855415b4c7566b)

CephContext: Add AssociatedSingletonObject to allow CephContext's singleton

If some objects associated to CephContext want to create a singleton object,
it can inherit AssociatedSingletonObject and implement destruction to get notified.

Signed-off-by: Haomai Wang <haomaiwang@gmail.com>
(cherry picked from commit 7fed5dee4f96a83d1d6914f6fc0895bba2d15b99)

mon/OSDMonitor: require mon_allow_pool_delete = true to remove pools

This is a simple safety check.  Since we default to true it is currently
opt-in.

Backport: giant, firefly
Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 6b982e4cc00f9f201d7fbffa0282f8f3295f2309)

Conflicts:
     src/test/mon/osd-pool-create.sh

     Changed $CEPH_MON to 127.0.0.1 -- the CEPH_MON was introduced after
     firefly to allow tests to run in parallel. Back in firefly all tests
     use the same port because 127.0.0.1 was hardcoded. We can't
     conveniently backport all that's necessary for tests to run in
     parallel, therefore we keep the 127.0.0.1 hardcoded.

Conflicts:
src/test/mon/osd-pool-create.sh
            TEST_no_pool_delete() follows a different test than in master

Merge pull request #5051 from SUSE/wip-12016-firefly

osd: cache full mode still skips young objects

Reviewed-by: Sage Weil <sage@redhat.com>

Merge pull request #5307 from dachary/wip-release-notes-firefly

backport v0.80.10 release notes to firefly

For pgls OP, get/put budget on per list session basis, instead of per OP basis, which could lead to deadlock.

Signed-off-by: Guang Yang (yguang@yahoo-inc.com)
(cherry picked from commit 0f884fdb31a26f241401ce2a9329dc0f2c4eccc7)

Conflicts:
src/librados/IoCtxImpl.cc
            In firefly, return value of objecter->pg_read() is not assigned to c->tid.
src/osdc/Objecter.cc
src/osdc/Objecter.h
            There is no _op_submit_with_budget() function in firefly.
            There is no Objecter::_finish_op() function in firefly.
            In firefly, _take_op_budget() is called take_op_budget().

Merge pull request #4867 from SUSE/wip-11878-firefly

mon: Clock skew causes missing summary and confuses Calamari

Reviewed-by: Joao Eduardo Luis <joao@suse.de>

Always provide summary for non-healthy cluster.

This fixes a problem, wherein calamari does not provide
popup drill-downs for warnings or errors, should the summary
be missing.

Calamari gets health info from /api/v1/cluster/$FSID/health.
If the data here has a summary field, this summary is provided
in a popup window:

  /api/v1/cluster/$FSID/health is populated (ultimately) with
  status obtained via librados python bindings from the ceph
  cluster. In the case where there's clock skew, the summary
  field supplied by the ceph cluster is empty.

No summary field, no popup window with more health details.

Signed-off-by: Thorsten Behrens <tbehrens@suse.com>
(cherry picked from commit eaf6e0cf48488fe604d0ef0db164d44948d4e8d4)

Conflicts:
src/mon/Monitor.cc
            firefly needs ss

Merge pull request #5062 from SUSE/wip-11699-firefly

objecter: resend linger ops on split

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

doc: v0.80.10 changelog

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit b3a3e0e509eb9530994723ed9d68f85b8434a22d)

doc/release-notes: v0.80.10

Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 2a61b611987d169d5a501d77fbd545068550c28c)

Conflicts:
doc/release-notes.rst

doc/release-notes: update notes for v0.80.10

Signed-off-by: Loic Dachary <loic@dachary.org>
(cherry picked from commit 1231ae09d675f2ee9e76a04b91f771d73a2fb346)

Conflicts:
doc/release-notes.rst

Merge pull request #5129 from SUSE/wip-12202-firefly

mon: ceph fails to compile with boost 1.58

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5044 from SUSE/wip-12011-firefly

osd: PG stuck with remapped

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5039 from SUSE/wip-11961-firefly

osd: Cleanup boost optionals for boost 1.56

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #5037 from SUSE/wip-11875-firefly

tools: ceph-post-file fails on rhel7

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4771 from SUSE/wip-11759-firefly

tools: ceph-disk: support NVMe device partitions

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4636 from SUSE/wip-11143-firefly

tools:  src/ceph-disk : disk zap sgdisk invocation

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4635 from SUSE/wip-7387-firefly

common: Malformed JSON command output when non-ASCII strings are present

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4632 from SUSE/wip-11507-firefly

osd: object creation by write cannot use an offset on an erasure coded pool

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4630 from SUSE/wip-10983-firefly

tools: use a new disk as journal disk,ceph-disk prepare fail

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4597 from xinxinsh/wip-11026-firefly

objecter: can get stuck in redirect loop if osdmap epoch == last_force_op_resend

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4582 from SUSE/wip-11543-firefly

tools: ceph_argparse_flag has no regular 3rd parameter

Reviewed-by: Loic Dachary <ldachary@redhat.com>

Merge pull request #4633 from SUSE/wip-11535-firefly

Client admin socket leaks file descriptors

Reviewed-by: Josh Durgin <jdurgin@redhat.com>

os/chain_xattr: handle read on chnk-aligned xattr

If we wrote an xattr that was a multiple of a chunk, we will try to read
the next chunk and get ENODATA.  If that happens bail out of the loop and
assume we've read the whole thing.

Backport: hammer, firefly
Signed-off-by: Sage Weil <sage@redhat.com>
(cherry picked from commit 8614dcebf384b74b283cd352001a65fa26ba925c)

Conflicts:
    src/test/objectstore/chain_xattr.cc
        added #define FILENAME

rgw: error out if frontend did not send all data

Fixes: #11851
The civetweb mg_write() doesn't return error when it can't flush all data
to the user, it just sends the total number of bytes written. Modified the
client io to return total number of bytes and return an error if didn't
send anything.

Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
(cherry picked from commit daa679c3dd3770a6d5421e2cc9a36924f4550439)

Unconditionally chown rados log file.

This fixes bnc#905047 (in a somewhat ad-hoc way). Sadly the log
file gets created from several places, so its existence does not
mean init-radosgw had actually run.

Signed-off-by: Thorsten Behrens <tbehrens@suse.com>
(cherry picked from commit 41611ea27a42d6221c418613d24dce34d988f2e6)

ceph.spec.in: use _udevrulesdir to eliminate conditionals

The conditionals governing where 50-rbd.rules is installed were not doing the
right thing on SUSE distros.

Start using the %_udevrulesdir RPM macro, while taking care that it is defined
and set to the right value. Use it to eliminate some conditionals around other
udev rules files as well.

http://tracker.ceph.com/issues/12166 Fixes: #12166

Signed-off-by: Nathan Cutler <ncutler@suse.com>
(cherry picked from commit 8aa758ee7a14e93ded1d55b4aca7a54aa47d7174)

Conflicts:
ceph.spec.in
            master spec file has diverged significantly from firefly

ceph.spec.in: do not run fdupes, even on SLE/openSUSE

In openSUSE there is a policy to use %fdupes in the spec file if RPMLINT
complains about duplicate files wasting space in the filesystem.

However, RPMLINT is not so complaining, so drop fdupes.

http://tracker.ceph.com/issues/12301 Fixes: #12301

Signed-off-by: Nathan Cutler <ncutler@suse.com>
(cherry picked from commit 53072b9019caf72e0313b2804ea174237ed7da33)