build: Link TBB into io_exerciser if present

libstdc++ uses TBB to implement the execution library if it is
available. If it's not present, we get a serial backend.

Currently, we aren't getting link errors in most cases because the
reference is optimized out, but when compiling with `-O0`, we hit a
missing symbol.

If we use more of the execution library, we'll reference TBB in ways
that don't optimize out.

As such, test if TBB is available. If so, link against it.

See https://gcc.gnu.org/onlinedocs/libstdc++/manual/status.html#status.iso.2017
for more information.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

Merge pull request #62597 from soumyakoduri/wip-skoduri-restore-doc

rgw: Updating doc for cloud-restore feature

Merge pull request #62524 from ivancich/wip-fix-gap-list-tool

rgw: fix bug with rgw-gap-list

Reviewed-by: Jesse Williamson <jfw@ibm.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #56499 from adamemerson/wip-rgw-close-datalog-write-hole

rgw: Close datalog write hole

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

Merge pull request #62583 from phlogistonjohn/jjm-refix-setsebool

qa/cephadm/smb: run setsebool with sudo

Reviewed-by: Adam King <adking@redhat.com>

Merge pull request #62639 from phlogistonjohn/jjm-doc-note-clustering-opt

doc/mgr: add a warning about the smb clustering option & placement

Reviewed-by: Adam King <adking@redhat.com>
Reviewed-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>
Reviewed-by: Avan Thakkar <athakkar@redhat.com>
Reviewed-by: Sachin Prabhu <sp@spui.uk>

Merge pull request #62604 from jzhu116-bloomberg/wip-70746

rgw: keep the tails when copying object to itself

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

Merge pull request #62325 from Matan-B/wip-matanb-with-crimson

src: sed -i 's/WITH_SEASTAR/WITH_CRIMSON/'

Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>

Merge pull request #61804 from aainscow/erasure_code_chunks

osd: Modify erasure code plugins ready for new EC

Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

Merge pull request #62233 from kamoltat/wip-ksirivad-ceph-s-leader-mon

mon: Add leader indicator && stretch mode to ceph -s output

Merge pull request #57806 from dongbula/wip-speed-oc-flush2

osdc/ObjectCacher: avoid io blocking for bufferheads exceed limit

Merge pull request #56612 from Matan-B/wip-matanb-remove-snapset-snaps

osd: cleanup Snapset::snaps

Merge pull request #62205 from leonidc/wip-leonidc-redeploy-fix

nvmeofgw* :do not allow failover for gws during redeploy (fast-reboot)

Merge pull request #62223 from aainscow/enable-switch-flag

osd:  Add ability to switch EC optimizations on

Reviewed-by: Bill Scales <bill_scales@uk.ibm.com>|
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Reviewed-by: Samuel Just <sjust@redhat.com>

Merge pull request #62641 from tchaikov/cmake-set_property

cmake: Fix warning suppression for googletest build

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

Merge pull request #62629 from rhcs-dashboard/cluster-mgr-role-fix

mgr/dashboard: add prometheus read permission to cluster_mgr role

Reviewed-by: Aashish Sharma <aasharma@redhat.com>

ceph.spec.in, install-deps: sed -i 's/seastar/crimson/'

Signed-off-by: Matan Breizman <mbreizma@redhat.com>

src: sed -i 's/WITH_SEASTAR/WITH_CRIMSON/'

Signed-off-by: Matan Breizman <mbreizma@redhat.com>

Merge pull request #62643 from zdover23/wip-doc-2025-04-03-cephadm-upgrade

doc/cephadm: s/confg/config/

Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

doc/cephadm: s/confg/config/

Correct a misspelled command.

Re: https://github.com/ceph/ceph/pull/62367#discussion_r2025832937

Signed-off-by: Zac Dover <zac.dover@proton.me>

cmake: Fix warning suppression for googletest build

In commit 27e9d563, we attempted to disable deprecated warnings when building
googletest, but the implementation contained two errors:

1. The `set_property()` call occurred before adding the target directory,
   making it impossible to set properties on non-existent objects.
2. The `-Wno-deprecated-declarations` flag was incorrectly passed as an
   `APPEND` argument instead of a `PROPERTY` argument.

This caused build failures with libstdc++-12 and newer Clang versions:

```
CMake Error at src/CMakeLists.txt:772 (set_property):
  set_property given invalid argument "-Wno-deprecated-declarations".
```

This commit fixes both issues by:
- Moving the `set_property()` call after `add_subdirectory()`
- Correctly passing the warning flag as a `PROPERTY` argument

Signed-off-by: Kefu Chai <tchaikov@gmail.com>

Merge pull request #62592 from adk3798/cephadm-rgw-only-bind-on-networks

mgr/cephadm: add only_bind_port_on_networks support for rgw

Reviewed-by: Kushal Deb <Kushal.Deb@ibm.com>
Reviewed-by: Shweta Bhosale <Shweta.Bhosale1@ibm.com>

Merge pull request #62362 from phlogistonjohn/jjm-mgr-smb-tweaks

mgr/smb: additional input and output controls

Reviewed-by: Avan Thakkar <athakkar@redhat.com>
Reviewed-by: Sachin Prabhu <sp@spui.uk>

Merge pull request #61461 from NitzanMordhai/wip-nitzan-ec-clay-profile-config-wrong-d

erasure-code: clay wrong d calculation

test: IO sequence exerciser enable ec optimisations

The io sequencer has been written to primarily test the new EC code.  This commit
turns that flag on by default.  It also provides a flag to disable the new
optimizations, which we expect to drop in the future.

Signed-off-by: Alex Ainscow <aainscow@uk.ibm.com>

osd: EC Switcher actually switches between two (identical) EC implementations

Signed-off-by: Alex Ainscow <aainscow@uk.ibm.com>

osd: Create a new interval when the optimization flag gets switch on.

This allows for the old EC code to cleanly shutdown using the on_change
notification. The new code can then start up cleanly.

Signed-off-by: Alex Ainscow <aainscow@uk.ibm.com>

osd: Add EC_OPTIMIZATIONS flag and add ability to switch it on.

This is the basic command. On its own, this does nothing functional, except set
a flag which can be queried.

Signed-off-by: Alex Ainscow <aainscow@uk.ibm.com>

doc/mgr: add a warning about the smb clustering option & placement

Add a warning to the docs highlighting that `clustering` is an advanced
option and setting it without out also setting a compatible placement
value may lead to unexpected behavior.

Signed-off-by: John Mulligan <jmulligan@redhat.com>

Merge pull request #61638 from jamiepryde/58052-release-note

PendingReleaseNotes: add note about changing default EC plugin from Jerasure to ISA-L

Merge pull request #62543 from rhcs-dashboard/smb-carbonize-tabs

mgr/dashboard: carbonize smb tabs

Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #62614 from rhcs-dashboard/fix-smb-placement

mgr/dashboard: fix smb cluster placement

Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #62294 from rzarzynski/wip-bypassable-ceph-assert

common, *: make the abort() in ceph_assert() bypassable

Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Reviewed-by: Adam Kupczyk <akupczyk@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>
Reviewed-by: Matan Breizman <mbreizma@redhat.com>

Merge pull request #55696 from bill-scales/main

common/mempool.cc: Improve performance of sharding

Reviewed-by: Adam Kupczyk <akupczyk@redhat.com>
Reviewed-by: Jose Juan Palacios-Perez <perezjos@uk.ibm.com>
Reviewed-by: John Agombar <agombar@uk.ibm.com>
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>

Merge pull request #58652 from adk3798/mgr-rgw-create-ec-data-pool

mgr/rgw: add ability for rgw realm bootstrap to create pools

Reviewed-by: Kushal Deb <Kushal.Deb@ibm.com>

Merge pull request #62611 from rhcs-dashboard/ec-fixes

mgr/dashboard: add enhancements for pool rulest

Reviewed-by: Aashish Sharma <aasharma@redhat.com>

Merge pull request #60614 from adk3798/host-maintenance-exit-refresh-daemons

mgr/cephadm: trigger host metadata refresh when taking host out of maintenance

Reviewed-by: Shweta Bhosale <Shweta.Bhosale1@ibm.com>

rgw: remove keep_tail from RGWObjState

Signed-off-by: Jane Zhu <jzhu116@bloomberg.net>

mgr/dashboard: carbonize smb tabs

also removes share tab as was the only one in favour of title and description

Fixes: https://tracker.ceph.com/issues/70695
Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>

Merge pull request #62622 from tchaikov/cmake-disable-deprecated-warning

cmake: disable deprecated warning when building googletest

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

mgr/dashboard: add prometheus read permission to cluster_mgr role

Fixes: https://tracker.ceph.com/issues/70768
Signed-off-by: Nizamudeen A <nia@redhat.com>

mgr/dashboard: fix smb cluster placement

Fixes: https://tracker.ceph.com/issues/70757
Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>

mgr/dashboard: add enhancements for pool rulest

On EC pool:
- Use host count instead of device count for host crush-failure-domain
- Host warning for k+m+1
On replicated:
- Set 'All devices' as default

Fixes: https://tracker.ceph.com/issues/70764
Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>

Merge pull request #62246 from aclamk/wip-aclamk-write-v2-tune-down-unused

os/bluestore/write-v2: Tune down usage of 'unused'

Merge pull request #62457 from rhcs-dashboard/nvmeof-upgrade-bug

mgr/dashboard: upgrading nvmeof doesn't update configuration

Merge pull request #62609 from aainscow/io_exerciser_restrict

test/common: EC Optimizations: rados io sequencer exerciser extensions

Reviewed-by: Bill Scales <bill_scales@uk.ibm.com>
Reviewed-by: Jonathan Bailey <jonathan.bailey1@ibm.com>

Merge pull request #61696 from MaxKellermann/osd_includes

osd: include cleanup

Reviewed-by: Matan Breizman <mbreizma@redhat.com>

Merge pull request #62542 from rhcs-dashboard/smb-service-section

mgr/dashboard: add separation for service section in smb cluster form

Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #62541 from rhcs-dashboard/fix-scroll

mgr/dashboard: fix domain settings modal scrollable

Reviewed-by: Nizamudeen A <nia@redhat.com>

PendingReleaseNotes: add note to change default EC plugin from Jerasure to ISA-L

Signed-off-by: Jamie Pryde <jamiepry@uk.ibm.com>

Merge pull request #62047 from rhcs-dashboard/fix-70226-main

monitoring: Fix OSDs panel in host-details grafana dashboard

Reviewed-by: Nizamudeen A <nia@redhat.com>

Merge pull request #62598 from rhcs-dashboard/nvmf-subsys-ha-default

mgr/dashboard: enable ha by default on subsystem POST API

cmake: disable deprecated warning when building googletest

In commit 986f6918, we updated the googletest submodule to silence a
CMake warning. However, this change broke the build due to a Clang issue
where it warns about deprecated declarations even from system headers
(see https://github.com/llvm/llvm-project/issues/76515). Since we use
`-Werror` in our compile options, these warnings become errors and fail
the build.

This change detects if we're affected by the specific combination of
compiler, standard library, and compile options that triggers this
issue. It then conditionally disables the `-Wdeprecated-declarations`
warning when building googletest to resolve build failures caused by
deprecated functions like `get_temporary_buffer<>` in the standard
library.

The build failure looks like:
```
FAILED: src/googletest/googletest/CMakeFiles/gtest.dir/src/gtest-all.cc.o
...
In file included from /home/jenkins-build/build/workspace/ceph-pull-requests/src/googletest/googletest/src/gtest-all.cc:38:
In file included from /home/jenkins-build/build/workspace/ceph-pull-requests/src/googletest/googletest/include/gtest/gtest.h:55:
In file included from /usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/memory:66:
/usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/stl_tempbuf.h:263:8: error: 'get_temporary_buffer<testing::TestInfo *>' is deprecated [-Werror,-Wdeprecated-declarations]
  263 |                 std::get_temporary_buffer<value_type>(_M_original_len));
      |                      ^
/usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/stl_algo.h:4996:15: note: in instantiation of member function 'std::_Temporary_buffer<__gnu_cxx::__normal_iterator<testing::TestInfo **, std::vector<testing::TestInfo *>>, testing::TestInfo *>::_Temporary_buffer' requested here
 4996 |       _TmpBuf __buf(__first, (__last - __first + 1) / 2);
      |               ^
/usr/lib/gcc/x86_64-linux-gnu/12/../../../../include/c++/12/bits/stl_algo.h:5070:23: note: in instantiation of function template specialization 'std::__stable_sort<__gnu_cxx::__normal_iterator<testing::TestInfo **, std::vector<testing::TestInfo *>>, __gnu_cxx::__ops::_Iter_comp_iter<(lambda at /home/jenkins-build/build/workspace/ceph-pull-requests/src/googletest/googletest/src/gtest.cc:3024:20)>>' requested here
 5070 |       _GLIBCXX_STD_A::__stable_sort(__first, __last,
      |                       ^
/home/jenkins-build/build/workspace/ceph-pull-requests/src/googletest/googletest/src/gtest.cc:3023:8: note: in instantiation of function template specializatio
```

Signed-off-by: Kefu Chai <tchaikov@gmail.com>

Merge pull request #62617 from anthonyeleven/pushme-pullme

doc/rados/configuration: Correct admonition in ceph-conf.rst

Reviewed-by: Zac Dover <zac.dover@proton.me>

doc/rgw: Updating doc for cloud-restore feature

Signed-off-by: Soumya Koduri <skoduri@redhat.com>

doc/rados/configuration: Correct admonition in ceph-conf.rst

Signed-off-by: Anthony D'Atri <anthonyeleven@users.noreply.github.com>

Merge pull request #62615 from zdover23/wip-doc-2025-04-02-pr-62613-followup

doc/dev/release-process: followup on PR#62610

doc/dev/release-process: followup on PR#62610

Make minor usage and markup changes to doc/dev/release-process.rst to
follow up on Dan Mick's contribution in https://github.com/ceph/ceph/pull/62610.

Signed-off-by: Zac Dover <zac.dover@proton.me>

rgw: keep the tails when copying object to itself

Signed-off-by: Jane Zhu <jzhu116@bloomberg.net>

Merge pull request #62610 from ceph/release-process-container-build

doc/dev/release-process.rst: document new Jenkins job for containers

doc/dev/release-process.rst: document new Jenkins job for containers

The manual instructions for building release containers were complex
and required access to an arm64 host; that was just too much,
so I've created a Jenkins job to build all of the prerelease
artifacts (see https://github.com/ceph/ceph-build/pull/2342).
Change the documentation to reflect that.

Signed-off-by: Dan Mick <dan.mick@redhat.com>

mgr/cephadm: trigger host metadata refrsh when taking host out of maintenance

If a host was in maintenance mode, we should make sure to check
all of the host information again once it's taken out even if
it hasn't been long enough for the daemon/device/facts etc. cache
to timeout. Some users may make changes to things like the node's
hardware or OS during maintenance mode that we should make sure to
pick up as soon as possible

Signed-off-by: Adam King <adking@redhat.com>

mgr/cephadm: add func to invalidate all host metadata and kick serve loop

We already had one of these specifically for daemons. This commit
just adds the same but for all types of host metadata. Currently
the intended use is to call this when taking a host out of maintenace
mode as it's not just the daemons that could have been modified
while the host was in maintenance mode.

Signed-off-by: Adam King <adking@redhat.com>

mgr/rgw: add ability for rgw realm bootstrap to create pools

Specifically by setting the data_pool_attributes field in
the RGW spec. For example

service_type: rgw
service_id: my_realm.my_zone
placement:
  count: 2
spec:
  data_pool_attributes:
    type: ec
    k: 6
    m: 2
  rgw_realm: my_realm
  rgw_zone: my_zone
  rgw_zonegroup: my_zonegroup

would create the my_zone.rgw.buckets.data pool as an
ec pool with a new ec porfile with k=8 and m=2. Currently
the ec profile fields this spec will make use of
are only "k", "m", "pg_num" and "crush-device-class".
If other attributes are set, or if the pool type is
"replicated" the key value pairs are assumed to
be ones that can be passed to the "ceph osd pool create"
command. The other pools for the rgw zone (e.g. buckets
index pool) are all made as replicated pool with
defaults settings

Signed-off-by: Adam King <adking@redhat.com>

test/common: rados io sequencer exerciser extensions

1. Add miscompare message containing object ID.
2. Fix compiler warning due to strangely placed assert.
3. Add barriers following error injects to insure they
   are in place before IO.
4. Do not, by default, test EC profiles that are not
   known to be good for EC optimisations.
5. Add "allow_unstable_pool_configs" to override above.

Signed-off-by: Alex Ainscow <aainscow@uk.ibm.com>

Merge pull request #62284 from yuvalif/wip-yuval-70086

rgw/logging: use bucket policy for logging

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

Merge pull request #62288 from athanatos/sjust/wip-crimson-recovery-69412-replica

crimson: take read lock on replica during final push commit

Reviewed-by: Matan Breizman <mbreizma@redhat.com>

neorados: Suppressions for Boost.asio

Signed-off-by: Adam Emerson <aemerson@redhat.com>

qa/rgw: Suppress valgrind error from unclean shutdown

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: Watch/Notify memory leak fix maybe!

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw: Add recovery semaphore commands to radosgw-admin

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Decrement with grace period

This guards against an excess decrement in the sequence:

RGW_a: Fetch sem_set (see key 'foo')
RGW_b: run renew_entry ('foo' is no longer in `cur_cycle`)
RGW_a: notify (does not see 'foo' in response)
RGW_a: Decrements 'foo'

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Don't read the entire sem_set per shard

Process shards piecewise. Send each set of shards in a window to the
other RGWs and have them acknowledge having it or not.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

clang14: Clang14 broken on lambda captures of structured bindings

This works fine on GCC11, which is the actual target we use for
builds on Shaman.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Make add_entry a stackful coroutine

Since, outside of testing, it's only called from stackful coroutines,
for now.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Test semaphores/recovery

Also tests renew_entry, backend, and list.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/admin: Don't start background tasks in radosgw-admin

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Orderly shutdown

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Parallelize recovery

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Semaphores and Recovery

Increment in add_entry, decrement in renew_entry, and recover on
startup.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: de/serialize BucketShards from/to a string

So we can reconstitute bucket shards from listings during recovery.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: add_entry() failures are fatal, actually

Except in one particular case.

This is necessary to close the data replication write hole, so that if
we fail to write to the semaphore set, the client will receive an
error.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Remove cls_fifo_legacy

Signed-off-by: Adam Emerson <aemerson@redhat.com>

test/rgw_multi: Fix unbound variable in log message

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw/multisite/datalog: C++20 Coroutine Refactor

Rewrite all of `RGWDataChangesLog` and supporting classes to use
non-blocking, C++20 coroutines.

Make interfaces for `optional_yield` and `librados::AioCompletion`.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Get rid of unneeded set_zero

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

rgw/multisite/datalog: Take driver, not handle.

Since we're plumbing the driver around and it's where our neorados
handle lives.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

common/async: add spawn_group template for fork-join parallelism

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

rgw: Add neorados functions to RGWObjVersionTracker

Signed-off-by: Adam Emerson <aemerson@redhat.com>

rgw: Add neorados utilities

Signed-off-by: Adam Emerson <aemerson@redhat.com>

cls: Add sem_set class

This stores a collection of key-based semaphores, intended to serve as
a backend to the project to close the datalog write-hole.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

gcc11: Rewrite some chunks of FIFO

Revert this commit when we are at last free of gcc11.

Signed-off-by: Adam Emerson <aemerson@redhat.com>

neorados/cls: Client for FIFO objclass

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

gcc11: Temporary change to neorados CLS log

Signed-off-by: Adam Emerson <aemerson@redhat.com>

neorados/cls: Client for log objclass

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

cls/log: C++ namespaces exist

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

cls/log: Switch from utime_t to ceph::real_time

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

cls/log: Add non-default constructors to cls_log_entry

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

cls/log: Switch from std::list to std::vector

We should not be using std::list everywhere, and this is an excellent
time to switch.

Signed-off-by: Adam C. Emerson <aemerson@redhat.com>

gcc11: Pull out exec from cls version

Signed-off-by: Adam Emerson <aemerson@redhat.com>