os/test/store_test: fix legacy bluestore cache settings application

Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>

Merge pull request #11894 from xiexingguo/xxg-wip-bluestore-1110

os/bluestore: speed up omap-key generation for same onode(the read paths)

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

Merge pull request #11893 from xiexingguo/xxg-wip-bluestore-fix-clear

os/bluestore: fix potential memory leak

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

Merge pull request #11892 from tchaikov/wip-move-context

Context: std::move the callback param in FunctionContext's ctor

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

Merge pull request #11889 from trociny/wip-vstart-bashism

vstart.sh: fix bashism in the script

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

Merge pull request #11601 from ceph/wip-ms-async-keepalive2

msg/async/AsyncConnection: dispatch write handler on keepalive2

Reviewed-by: Haomai Wang <haomai@xsky.com>

Merge pull request #11633 from ddiss/tcc17684_rados_segv_without_pool

rados: fix segfaults when run without --pool

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

Merge pull request #11665 from wido/mon-prevent-pool-delete

mon: Do not allow pools to be deleted by default

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

Merge pull request #11795 from runsisi/wip-fix-osd-noin

mon/OSDMonitor: newly created osd should not be wrongly marked in

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

Merge pull request #10049 from sponce/wip-striper-async-rm-master

librados: improvements async IO in librados and libradosstriper

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

Context: std::move the callback param in FunctionContext's ctor

for better performance.

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

Merge pull request #11883 from jcsp/wip-mgr-functioncontext

mgr: remove unnecessary C_StdFunction

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

os/bluestore: fix potential memory leak

Call to extent_map.clear_and_dispose() will be noop after extent_map.clear()
hence causing memory leak.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

vstart.sh: fix bashism in the script

Signed-off-by: Mykola Golub <mgolub@mirantis.com>

os/bluestore: speed up omap-key generation for same onode(the read paths)

In https://github.com/ceph/ceph/pull/11807 we only modify the write paths,
this patch fixes the read parts too, which are more time-sensitive.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

Merge pull request #11886 from stiopaa1/mds_mantle_includeCorrectHeaders

mds/Mantle.h: include correct header files

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>

Merge pull request #11683 from wjwithagen/wip-wjw-freebsd-vstart

src/vstart.sh: Only execute btrfs if it is available

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

Merge pull request #11702 from wjwithagen/wip-wjw-freebsd-cephtool-test

qa/workunits/cephtool/test.sh: FreeBSD has no distro.

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

Merge pull request #11885 from cbodley/wip-python-rgw-deb

deb: update python-rgw dependencies to librgw2

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

mds/Mantle.h: include correct header files

Signed-off-by: Michal Jarzabek <stiopa@gmail.com>

deb: update python-rgw dependencies to librgw2

Signed-off-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #11881 from linuxbox2/rgw-file-fsid

rgw_file: fix spurious mount entries w/Linux NFS client
verified that failed check is in osd-scrub-repair.sh

Merge pull request #11882 from yehudasa/wip-yehuda-spdk

spdk: a few fixes

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

mgr: remove unnecessary C_StdFunction

FunctionContext works fine and already existed.

Signed-off-by: John Spray <john.spray@redhat.com>

Merge pull request #11811 from tanghaodong25/reap-ioc-wip

os/bluestore: reap ioc when stopping aio_thread.

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

rgw_file: fix spurious mount entries w/Linux NFS client

This change fixes the initialization of root file handles to have
their correct fs instance.  Previously, the root file handle had
fs_inst-1.  The apparent change in fsid in a subordinate directory
triggers the client's behavior for new filesystem mounts, which
we don't want.

Fixes: http://tracker.ceph.com/issues/17850
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #11832 from ceph/wip-python-rgw

deb: add python-rgw packages

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

debian: add python-rgw package

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

Merge pull request #11845 from xiexingguo/xxg-wip-bs-1109

os/bluestore: remove duplicated namespace of tx state

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

Merge pull request #11859 from ceph/wip-py-rgw

doc: add python-rgw doc

Reviewed-by: Haomai Wang <haomai@xsky.com>

Merge pull request #11823 from xiexingguo/xxg-wip-bluestore-1108

os/bluestore: refactor _do_write(); move initializaiton of csum out of loop

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>

Merge pull request #10748 from yuyuyu101/wip-dpdk8

msg/async: DPDKStack as AsyncMessenger backend

doc: add doc for pybind of rgw

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

pybind: rename rgw_file.pyx to rgw.pyx

so the init function offered by rgw python extension is named "initrgw"
instead of "initrgw_file". and "initrgw" is the expected name by python
intepreter when loading an extension.

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

Merge pull request #11847 from yuyuyu101/wip-fix-rgw-spec

ceph.spec: add pybind rgwfile

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

ceph.spec: add pybind rgwfile

Signed-off-by: Haomai Wang <haomai@xsky.com>

os/bluestore: remove duplicated namespace of tx state

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

test/CMakeLists: only enable ceph_test_userspace when dpdk is true

Signed-off-by: Haomai Wang <haomai@xsky.com>

msg/async/dpdk: change error to warning

Signed-off-by: Haomai Wang <haomai@xsky.com>

update dpdk submodule to ceph/dpdk

Signed-off-by: Haomai Wang <haomai@xsky.com>

msg/async/Stack: allow to setup dpdkstack

Signed-off-by: Haomai Wang <haomai@xsky.com>

test_async_networkstack: fix unmatched create/delete file event

Signed-off-by: Haomai Wang <haomai@xsky.com>

test/test_async_networkstack: add dpdk backend test

Signed-off-by: Haomai Wang <haomai@xsky.com>

test: add userspace_event

Signed-off-by: Haomai Wang <haomai@xsky.com>

Event: don't wakeup dpdk driver since it's spinning on cpu

Signed-off-by: Haomai Wang <haomai@xsky.com>

msg/async/Event: add DPDKDriver type to allow userspace event

Like epoll, kqueue, select, dpdk backend also need to allow notify EventCenter
when a connection is ready to read/write. So we need to let DPDKStack can kick
EventCenter. DPDKDriver which introduced UserspaceEvent to simulate epoll
event achieve the goal.

Signed-off-by: Haomai Wang <haomai@xsky.com>

msg/async: add dpdk stack plugin

Signed-off-by: Haomai Wang <haomai@xsky.com>

buffer: add new ptr type which destruct with passed deleter

deleter is a wrapper for lambda which can used to free resource associated
with object

Signed-off-by: Haomai Wang <haomai@xsky.com>

msg/async/Event: add Poller to allow event thread spin on Poller::poll

Signed-off-by: Haomai Wang <haomai@xsky.com>

Event: Add get_driver to return driver backend

Signed-off-by: Haomai Wang <haomai@xsky.com>

Throttler: Add reset support

Signed-off-by: Haomai Wang <haomai@xsky.com>

os/bluestore: move selectiong of csum/crr out of for loop

We don't have to trace the csum/compression_required_ration option changing
at such a small granularity, which costs too much.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

os/bluestore: refactor _do_write()

By moving garbage collection logic into a separate method.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

Merge pull request #11681 from stiopaa1/mds_mdstableclient_addConsttoMemFun

mds/MDSTableClient.h: add const to member funct

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

Merge pull request #11818 from liewegas/wip-bluestore-aio-stalls3

os/bluestore: optional debug mode to identify aio stalls

Merge pull request #11826 from ovh/wip-doc-update-xattr-config-documentation

doc: Update filestore xattr config documentation.

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

Merge pull request #11831 from ceph/wip-dang-rgw-pybind

RGW Python bindings - use explicit array

RGW Python bindings - use explicit array

Older Cython doesn't to implicit simple arrays, so use an explicit array
for stat output.

Revert change removing stat

Signed-off-by: Daniel Gryniewicz <dang@redhat.com>

Merge pull request #11816 from cbodley/wip-17812

rgw: delete entries_index in RGWFetchAllMetaCR

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

os/bluestore: reap ioc when stopping aio_thread.

there is possibillity that reap_ioc don't excute when stopping
aio_thread. so add reap_ioc after aio_thread is stopped.

Signed-off-by: Haodong Tang <haodong.tang@intel.com>

Merge pull request #11817 from linuxbox2/wip-boost-3

boost: embedded

Merge pull request #11825 from ceph/wip-11820

pybind: fix build failure of rgwfile binding

Reviewed-by: Haomai Wang <haomai@xsky.com>

doc: Update filestore xattr config documentation.

Signed-off-by: Bartłomiej Święcki <bartlomiej.swiecki@corp.ovh.com>

pybind: fix build failure of rgwfile binding

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

Merge pull request #9624 from gaurav36/coverity-1355603-dead-code

coverity fix: fixing few coverity issue

Merge pull request #11798 from stiopaa1/mds_fsmap_preventcopyinginloop

mds/FSMap.cc: prevent unneeded copy of map entry

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

Merge pull request #11802 from stiopaa1/mds_fsmap_constReference

mds/FSMap.h: add const and reference

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

Merge pull request #11819 from stiopaa1/mds_migrator_addConstTomemFun

mds/Migrator.h: add const to member functions

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

Merge pull request #11807 from xiexingguo/xxg-wip-bs-1107

os/bluestore: speed up omap-key generation for same onode

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

os/bluestore: speed up omap-key generation for same onode

This is useful if we have to batch-process omap kvs.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>

spdk: cmake fixes

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

bluestore: trim serial number

remove any non-alpha character appended at the end

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

os/bluestore: null terminate readlink() result

readlink() does not guarantee null termination of
result.

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

Merge pull request #11754 from athanatos/wip-17789

osd/ReplicatedPG: do_update_log_missing: take the pg lock in the callback

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

os/bluestore/KernelDevice: debug support to detect stalled aios

This is additional bookkeeping to track the oldest aio in flight, so we
make it optional, and disabled by default.  It is useful for identifying
buggy kernel aio or bad devices.

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

os/fs: add aio_list_t type

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

os/fs: uninline aio methods

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

mds/Migrator.h: add const to member functions

Signed-off-by: Michal Jarzabek <stiopa@gmail.com>

boost:  pull up boost to enable boost_python build

Pullup to just the merges on Sep 10 fixes an obvious compile
problem, but there are other relevant changes.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

Adapt for ceph-manager

The ceph manager change added a separate Boost dependency
check, add it to the combined section.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

do_cmake.sh: set BOOST_J to $(nproc)

This lets us build boost with all processors.

Signed-off-by: Sage Weil <sage@redhat.com>
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

boost:  update boost to 8/29

Pull-up Boost to the commit which merged the upstream fix for
upstream Boost/iostreams bug #11860 (template arity warnings),
per discussion in core.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

cmake:  find and build bundled boost

Build Boost using the src/boost submodule, unless overridden
by -DWITH_SYSTEM_BOOST.

If -DBOOST_J=<n> is provided, builds Boost with <n> jobs.

Boost builds in the configured Build directory.

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

boost: add submodule at tag boost-1.61.0

This implements option #4 for external boost, based on upstream
discussion.  In option #4:

1. boost is added as a submodule
2. builds default to using the attached boost module
3. building against a system-provided boost is supported, but
   must be configured explicitly

Because all of the boost components are attached as nested
submodules in the upstream boost repository, neither the nested
submodules nor the root boost submodule have been cloned into
modules in github.com/ceph (acked by Sage).

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

Merge pull request #11647 from ceph/wip-jlayton-cephfs

libcephfs client API overhaul and update

Reviewed-by: John Spray <john.spray@redhat.com>

Merge pull request #11425 from weiqiaomiao/wqm-rgw-oplog

rgw: fix the field 'total_time'  of log entry in log show opt

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

Merge pull request #8712 from diwakargoel/master

doc: Correcting the sample python tempurl generation script.

Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>

Merge pull request #11648 from cbodley/wip-17698

rgw: fix for bucket delete racing with mdlog sync

Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>

Merge pull request #11598 from cbodley/wip-rgw-bucket-sync

rgw multisite: move lease up to RunBucketSync instead of child crs

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

rgw: delete entries_index in RGWFetchAllMetaCR

use a unique_ptr to ensure the entries_index is freed on destruction

Fixes: http://tracker.ceph.com/issues/17812
Signed-off-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #11715 from dzafman/wip-test-ec-eio

Fix racey test by setting noout flag (tracker 17757)

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

Merge pull request #11624 from yuyuyu101/wip-rgwfile-pybind

Wip rgwfile pybind

pybind/rgwfile: change description

Signed-off-by: Haomai Wang <haomai@xsky.com>

pybind/rgwfile: fixes hardcode way to find include dir

Signed-off-by: Haomai Wang <haomai@xsky.com>

AsyncConnection: dispatch write handler on keepalive2

...otherwise, if the connection is idle (i.e. no proper TAG_MSG
messages), keepalive ack will not get sent until some point in the
future.

Fixes: #17664
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

AsyncConnection: rename _send_keepalive_or_ack -> _append_keepalive_or_ack

Commit 48d929c4f056 ("msg: async: improve _send_keepalive_or_ack()
a little") removed _try_send() call.  _try_send(false) was a no-op by
then, but still - rename appropriately.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

AsyncConnection: eliminate reader vs writer ack race in process()

process() will potentially handle multiple messages, yet we are
scheduling the writer to write out acks only once, after the first
message is processed.  This effectively races the reader with the
writer and leaves the window for exiting the process() loop with
pending acks (ack_left > 0 but no kicked writer).

Close this window by moving dispatch_event_external() out of the loop.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

pybind/rgwfile: raise Exception when failing to umount

Signed-off-by: Haomai Wang <haomai@xsky.com>

pybind/rgwfile: change FileHandler to FileHandle

Signed-off-by: Haomai Wang <haomai@xsky.com>

Merge pull request #11794 from rzarzynski/wip-rgw-17803

rgw: fix uncompressed object size deduction in RGWRados::copy_obj_data.

Reviewed-by: Casey Bodley <cbodley@redhat.com>