cls/rbd: changes needed to align rest with the proposed ones

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: proposed changes

Changes proposed by 'N Balachandran'

Signed-off-by: N Balachandran <nibalach@redhat.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: group replayer work in-progress changes

This commit is a WIP targeting stability of tests failures, it might be diluted
into various existing commits

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

qa/workunits/rbd: mirror group tests improvements

Signed-off-by: N Balachandran <nibalach@redhat.com>
Signed-off-by: John Agombar <agombar@uk.ibm.com>
Signed-off-by: Adam Lyon-Jones <adamlyon@uk.ibm.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: move the rename detection logic to snapshot GroupReplayer

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: use group_header object for resync flagging

Also move the resync checks to snapshot GroupReplayer

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: rollback to last good snapshot just while force promote

fix the group membership to match the rollback snapshot

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: fix the below bugs

* fix braces in the imageMap update_images_added & update_images_removed
* do not allow image add from non-primary
* `down+unknown` status shown on querying individual images which are part
  of group enabled for mirroring
* `mirror pool status` shows down+unknown status
* fix imageMap being overwritten when multiple images are enabled for mirroring
* fix misleading error msg when getting status of a non-mirror enabled group

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: prune snapshots added to the list on image_replayer shutdown

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: discover primary demote snapshot in group_replayer

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: remove group snaps on primary at snapshot creation

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: Independent GroupReplayer

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

qa/workunits/rbd: mirror group functional tests

Also sets the RBD_MIRROR_INSTANCES to 1 to avoid any deviations.

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: implement group resync functionality

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: support mirroring regular non-mirror group snapshots

This commit also enable deep copying `.group` snaps.

$ rbd --cluster site-a snap ls pool1/test_image1 --all --debug-rbd=0
SNAPID  NAME                                                                                       SIZE     PROTECTED  TIMESTAMP                 NAMESPACE
     4  .group.2_10416b8b4567_104d6b8b4567                                                         128 MiB             Mon Apr 22 17:07:57 2024  group (pool1/test_group@group_snap1)
     7  .group.2_10416b8b4567_104f6b8b4567                                                         128 MiB             Mon Apr 22 17:07:59 2024  group (pool1/test_group@group_snap2)
     8  .mirror.primary.72855be4-1ffb-4094-8426-fb1d5f082c21.71b62dad-f515-43cf-91b2-bf1225c5a0fc  128 MiB             Mon Apr 22 17:08:03 2024  mirror (primary peer_uuids:[5f9ea7aa-fa5b-4d2e-a098-3afe006361aa])

$ rbd --cluster site-b snap ls pool1/test_image1 --all --debug-rbd=0
SNAPID  NAME                                                                                           SIZE     PROTECTED  TIMESTAMP                 NAMESPACE
     5  .group.2_10416b8b4567_104d6b8b4567                                                             128 MiB             Mon Apr 22 17:08:07 2024  group
     7  .group.2_10416b8b4567_104f6b8b4567                                                             128 MiB             Mon Apr 22 17:08:08 2024  group
     8  .mirror.non_primary.72855be4-1ffb-4094-8426-fb1d5f082c21.64488da7-3350-4d04-8e4d-90047182e004  128 MiB             Mon Apr 22 17:08:09 2024  mirror (non-primary peer_uuids:[] 6d963873-6679-483b-b05a-8bf2536f4fdf:8 copied)

$ rbd --cluster site-a group snap ls pool1/test_group --debug-rbd=0
NAME                                 STATUS
group_snap1                              ok
group_snap2                              ok
.mirror.2_10416b8b4567_10536b8b4567      ok

$ rbd --cluster site-b group snap ls pool1/test_group --debug-rbd=0
NAME                                 STATUS
.mirror.2_10376b8b4567_1037327b23c6      ok
group_snap1                              ok
group_snap2                              ok
.mirror.2_10416b8b4567_10536b8b4567      ok

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: support rename with group mirroring

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: add undo code, exclusive locking and quiescing

* add essential logic to undo partially succeeded API's like, group promote,
  group demote, group enable, group disable, group image add and
  group image remove
* add exclusive locking and quiescing with-in all the required group API's
* adress code duplication and optimization with in the group API's

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: address partial group snapshots case

Make sure group snapshots doesn't get copied to secondary if the group snapshot
is incomplete on primary. On creation time of a group snapshots on primary,
make sure to delete the previous snapshot in case it is incomplete.

Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: request group snapshot creation when creating group image snapshot

It makes the group image replayers to synchronize and to have the
group snapshot created.

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: unlink group snapshot when pruning non-primary snapshot

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: initial group replayer implementation

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: hook GroupReplayer

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: make pool watcher also refresh groups

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: introduce generalized mirror entity (type, global_id, size)

where type may be image or group. Make pool watcher and image
mapper use it currently for images. The group support is coming.

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd-mirror: remove dead code

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

rbd: add 'mirror group' commands

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: link group snapshot when creating non-primary snapshot

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: unlink group snapshot when removing mirror snapshot

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: allow to add image to group on creation

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: add mirror group API

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: don't send 'image updated' notifications

when enabling/disabling group mirroring

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: introduce 'group updated' mirroring watcher notification

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd: introduce 'group add/remove image' async requests

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

librbd/api: assume user namespace for group snapshot

when the snapshot is specified by name only

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: add method to unlink image snapshot from group snapshot

When no image snapshots are left the group snapshot will be
automatically removed.

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: make possible mirror_image_list filter out group images

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: add async versions of some group functions

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: add group_spec and group_snap_id to mirror snapshot

which are going to be used when creating a group mirror snapshot.

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: add mirror group types and methods

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: add group snapshot namespace

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

cls/rbd: rename GroupSnapshotNamespace to GroupImageSnapshotNamespace

Signed-off-by: Mykola Golub <mgolub@suse.com>
Signed-off-by: Prasanna Kumar Kalever <prasanna.kalever@redhat.com>

Merge pull request #62936 from cbodley/wip-doc-rgw-getobjattrs

doc/rgw: release note for GetObjectAttributes

Reviewed-by: Adam C. Emerson <aemerson@redhat.com>
Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>

Merge pull request #62845 from rhcs-dashboard/fix-path

mgr/dashboard: fix smb edit resources

Reviewed-by: Afreen Misbah <afreen@ibm.com>

Merge pull request #62715 from cbodley/wip-qa-rgw-no-gc

qa/rgw: run verify tests with garbage collection disabled

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

Merge pull request #62921 from idryomov/wip-71026

librbd: disallow "rbd trash mv" if image is in a group

Reviewed-by: Ramana Raja <rraja@redhat.com>

Merge pull request #62941 from MaxKellermann/mds_Locker__abort

mds/Locker: use ceph_abort_msg() instead of ceph_assert()

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge pull request #59248 from kamoltat/wip-ksirivad-improve-netsplit-warning

HealthMonitor: Add topology-aware netsplit detection and warning

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

mds/Locker: use ceph_abort_msg() instead of ceph_assert()

This ceph_assert() always fails, but depending on the configuration
value `ceph_assert_supresssions`, execution may continue, but the
`dir` variable is left uninitialized.  This leads to a compiler
warning:

 /home/jenkins-build/build/workspace/ceph-api/src/mds/Locker.cc:451:22: error: variable 'dir' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]

clang then suggests to nullptr-initialize the variable:

 /home/jenkins-build/build/workspace/ceph-api/src/mds/Locker.cc:447:11: note: initialize the variable 'dir' to silence this warning
   447 |         CDir *dir;
       |                  ^
       |                   = nullptr

This, however, is a very bad idea because all this does is suppress
the warning; it still crashes the process.

Since there's no recovery from this problem, let's switch to
ceph_abort_msg() which is [[noreturn]] and the compiler can deduce
that `dir` is always initialized when it's used.

Signed-off-by: Max Kellermann <max.kellermann@ionos.com>

Merge pull request #62693 from ronen-fr/wip-rf-iocnt

osd/scrub: performance counters for I/O performed by the scrubber

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

Merge pull request #62898 from nbalacha/wip-nbalacha-70963

rbd: display mirror state creating

Reviewed-by: Ramana Raja <rraja@redhat.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>

Merge pull request #60899 from clwluvw/curl-einval

rgw: handle EINVAL translation in forward_request

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

Merge pull request #62888 from clwluvw/neorados-fifotrim

neorados: relax fifo trim error for ENODATA

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

doc/rgw: release note for GetObjectAttributes

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

Merge pull request #62902 from cbodley/wip-70700-disable

cmake/common: temporarily remove decode_start_v_checker tests

Reviewed-by: Dan Mick <dmick@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>

Merge pull request #60227 from clwluvw/zonegroup-delbucket

rgw: skip empty check on non-owned buckets by zonegroup

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

Merge pull request #62738 from clwluvw/copy-obj-remote-zonegroup

rgw: dont store replication attrs on remote copy obj

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

osd/scrub: count scrub I/O

Implement I/O counting in the PGBackend::be_scan_list()
and relevant functions it calls.

Signed-off-by: Ronen Friedman <rfriedma@redhat.com>

Merge pull request #62699 from Matan-B/wip-matanb-crimson-ignore-abort-v2

crimson/common/errorator: rework aborts error handlers

Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>

Merge pull request #62556 from aainscow/ec_pr_and_prereqs

osd: Optimised EC

Reviewed-by: Radoslaw Zarzynski <rzarzynski@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>

rbd: display correct mirror state when creating

The mirror image state is set to MIRROR_IMAGE_STATE_CREATING
when the image is first created on the secondary, but was displayed
as "unknown" by the rbd info command. This has been fixed.

Fixes: https://tracker.ceph.com/issues/70963
Signed-off-by: N Balachandran <nithya.balachandran@ibm.com>

Merge pull request #62710 from bill-scales/ec_backfill

osd: EC Optimizations: Backfill changes for partial writes

Merge pull request #62725 from VallariAg/nvmeof-teuthology-fio

qa/suites/nvmeof: Fix thrasher and fio script

Merge pull request #60731 from joscollin/wip-B68954-check-headers-journal-recovery

cephfs-journal-tool: check the headers in dump file after journal recovery

Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Reviewed-by: Rishabh Dave <ridave@redhat.com>

Merge pull request #62914 from baum/ms_dispatch2_clean_up

src/nvmeof/NVMeofGwMonitorClient.cc: ms_dispatch2 clean up

Merge pull request #62696 from anthonyeleven/mgr-prom

doc/mgr: Improve prometheus.rst

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

Merge PR #62577 into main

* refs/pull/62577/head:
libcephfs_proxy: avoid libc buffering for logging

Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Anoop C S <anoopcs@cryptolab.net>

Merge branch 'main' into mgr-prom

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

qa: test 'journal import' recognizes invalid headers post journal recovery

Fixes: https://tracker.ceph.com/issues/68954
Signed-off-by: Jos Collin <jcollin@redhat.com>

cephfs-journal-tool: check the headers in dump file after journal recovery

Fixes: https://tracker.ceph.com/issues/68954
Signed-off-by: Jos Collin <jcollin@redhat.com>

doc/mgr: Improve prometheus.rst

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

Merge pull request #62911 from bluikko/doc-cleanup-radosgw

doc/radosgw: Fix indentation in admin.rst

Merge pull request #62896 from zdover23/wip-doc-2025-04-21-revert-62782-c4f0f8e

doc: Revert "doc/mgr: Promptify CLI commands and other formatting fixes"

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

doc/rados/operations/health-checks: Add MON_NETSPLIT Warning

Fixes: https://tracker.ceph.com/issues/67371
Signed-off-by: Kamoltat Sirivadhna <ksirivad@redhat.com>

HealthMonitor: Add topology-aware netsplit detection and warning

Problem:
Currently, Ceph cannot detect and report network partitions (netsplits)
between monitors in different topology locations in a consolidated way.
While stretch mode can handle partitions through monitor elections,
users lack visibility into the topology-level view of network
disconnections, making troubleshooting difficult.

Solution:
This implementation adds a hierarchical netsplit detection mechanism that:
- Uses DirectedGraph structure for netsplit detection
- Maps monitor disconnections to relevant CRUSH topology levels
- Aggregates individual disconnections into location-level reports when appropriate
- Detects complete location-level netsplits when ALL monitors between locations
  cannot communicate
- Reports specific topology locations experiencing complete communication failures
- Falls back to individual monitor-level reporting for partial disconnections
- Handles monitors with missing location data gracefully
- Leverages HealthMonitor::check_for_mon_down to receive a set of down monitors,
  efficiently avoiding false netsplit reports for monitors already known to be down
- Implements smart filtering that correctly excludes down monitors from location-based
  analysis, ensuring accurate netsplit reporting at both individual and topology levels

The implementation produces user-friendly health warnings:
1. For complete location netsplits: "Netsplit detected between dc1 and dc2"
2. For individual monitor disconnections: "Netsplit detected between mon.a and mon.d"

Performance considerations:
- Time complexity: O(m²) where m is the number of monitors
- Space complexity: O(m²) for connection tracking
- Practical impact is minimal as monitor count is typically small (3-7)

Fixes: https://tracker.ceph.com/issues/67371
Signed-off-by: Kamoltat Sirivadhna <ksirivad@redhat.com>
Conflicts:
src/mon/Elector.cc - Trivial Fix

Merge pull request #62416 from kamoltat/wip-ksirivad-fix-connection-score

librbd: disallow "rbd trash mv" if image is in a group

Removing an image that is a member of a group has always been
disallowed.  However, moving an image that is a member of a group to
trash is currently allowed and this is deceptive -- the only reason for
a user to move an image to trash should be the intent to remove it.

More importantly, group APIs operate in terms of image names -- there
are no corresponding variants that would operate in terms of image IDs.
For example, even though internally GroupImageSpec struct stores an
image ID, the public rbd_group_image_info_t struct insists on an image
name.  When rbd_group_image_list() encounters a trashed member image
(i.e. one that doesn't have a name), it just fails with ENOENT and no
listing gets produced at all until the offending image is restored from
trash.  Something like this can be very hard to debug for an average
user, so let's make rbd_trash_move() fail with EMLINK the same way as
rbd_remove() does in this scenario.

The one case where moving a member image to trash makes sense is live
migration where the source image gets trashed to be almost immediately
replaced by the destination image as part of preparing migration.

Fixes: https://tracker.ceph.com/issues/71026
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

pybind/rbd: add ImageMemberOfGroup exception

EMLINK is returned by rbd_remove() if the image is a member of a group.
Add a dedicated exception similar to ImageBusy or ImageHasSnapshots and
a test for it.

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

rbd: don't print "image will expire at" message when trash_move() fails

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

Merge pull request #61212 from rishabh-d-dave/mgr-vol-count-clones

mgr/vol: count number of ongoing clones in CloneProgressReporter...

Reviewed-by: Milind Changire <mchangir@redhat.com>

Merge pull request #62870 from MaxKellermann/mds_includes

mds: include cleanup

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

crimson: remove any assert_failure pre assert usages

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

crimosn/common/errorator: cleanup assert_failure pre_assert

Any usage should be replaced with a message that supports priniting the
error.

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

crimson/common/errorator: assert_failure to print error

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

crimson/osd: Verbose assert_all aborts

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

crimson/common/errorator: allow assert_all to accept c_str()

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

crimson/common/errorator: Cleanup assert_all pre_assert

Not used

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

crimson/common/errorator: cleanup ErrorT::handler call

call error_t::handle without decalring handler and invoking it later on

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

crimson/ertr: assert_all informs about error being handled that way

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

test/crimson/test_errorator: ignore assert_all

This came up during: https://tracker.ceph.com/issues/69406#note-25
Where an "assert_all" was called but didn't cause an abort.
Added "ignore_assert_all" to showcase this scenario along with any
other case where we are expected to abort.
The tests could be used to verify errorator's aborting behavior.

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

crimson/common/errorator: print abort message when possible

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

crimson/common/errorator: Always check exception type

We shouldn't bypass this check in the is_same_v<return_t, no_touch_error_marker>
case.

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

crimson/common/errorator: add TODO

There are few TODOs around errorator code which might be worth looking
into: https://tracker.ceph.com/issues/70875

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

crimson/common/errorator: introduce take_exception_from_future

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

crimson/common/errorator: move exception_comment

move the comment to where __cxa_exception_type is used
to keep handle() comments shorter.

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

crimson/common/errorator: fix skipped aborts

We should also invoke the errfunc (which aborts) when the
return type is no_touch_error_marker.
Added comments explaining:
* why it's forbidden to return void
* why std::is_same_v<return_t, no_touch_error_marker> is checked

Signed-off-by: Xuehan Xu <xuxuehan@qianxin.com>
Signed-off-by: Matan Breizman <mbreizma@redhat.com>

Merge pull request #62708 from rishabh-d-dave/vols-snap-path

mgr/vol: add command to get snapshot path

Reviewed-by: Venky Shankar <vshankar@redhat.com>

Merge pull request #62837 from athanatos/sjust/wip-crimson-stuck-backfilling

crimson: fix several bugs causing stuck backfills

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

Merge pull request #62619 from athanatos/sjust/wip-replica-read-crimson-mosdpct

crimson: add MOSDPGPCT support

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

src/nvmeof/NVMeofGwMonitorClient.cc: ms_dispatch2 clean up

- return ACKNOWLEDGED/HANDLED
- remove registration for unwanted keys

Signed-off-by: Alexander Indenbaum <aindenba@redhat.com>

Merge pull request #62899 from tchaikov/cmake-build-boost

cmake: Fix b2 build with postfixed compiler versions

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