s3:rpc_client: remove references to rpc_pipe_client from wsp_cli.c

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit dc6d3c96c33797163ac578675b255872deca9b98)

s3:rpc_client: remember rpc_pipe_client->print_username

This is ugly, but it's less ugly than doing it deep
inside of cli_spoolss.c and other code.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 344e45507e122137c357fc177bb024f09fff8b01)

s4:librpc: make all but dcerpc_pipe->binding_handle internal struct members

We could use a dcerpc_internal.h for struct dcecli_security and
struct dcecli_connection, but in struct dcerpc_pipe we still
expose binding_handle and changing that would require way too
much work for now...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 353af4b70559be1665df345655af43a61177c710)

s4:torture/rpc: make use of dcerpc_binding_handle_get_transport()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 382ece66063bd8ef465e989857394e6f412df6d3)

s4:tortore/rpc: make use of dcerpc_binding_get_abstract_syntax() and dcerpc_binding_get_flags()

We should not use p->syntax nor p->transfer_syntax...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 1713ad96a42e9245697a29d4899924cfbf636fc5)

s4:torture/rpc: make use of dcerpc_binding_handle_get_binding() and dcerpc_binding_get_flags()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 7ed9867fa82ae61a4522a724580900adce7e2bb5)

s4:libnet: make use of dcerpc_binding_handle_get_transport()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit a4b1845f4c5202afedd65b9d0db2866de75420a5)

s4:libnet: make use of dcerpc_binding_handle_get_binding()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit ad20c6656aa6798541b71b172534054c172d3aa5)

librpc/rpc: un-const dcerpc_default_transport_endpoint()

It returns an alloced string so it should not be marked as
const...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 9fe85a93539adb73ae59247d044746073915684a)

librpc/rpc: add dcerpc_binding_handle_get_transport() helper function

This will allow callers to avoid the dcerpc_binding_handle_get_binding()
dcerpc_binding_get_transport() dance...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 0e751215b349e10cc796065d6d01de35f52b41fd)

librpc/rpc: add dcerpc_binding_handle_get_binding()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit a038e7d8cf7a93ecf9a3e0c906f0d0db81ecee54)

s3:rpc_client: implement rpccli_bh_get_binding()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit fffc6e0f37a2d6a6612e17ad6fa777ae6a8eb3b6)

s4:librpc/rpc: implement dcerpc_bh_get_binding()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 0ef5d79eb6feefa2b7c5821ced98773d6d8fdd3a)

s4:librpc/rpc: call dcerpc_binding_set_abstract_syntax after bind or alter context

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit e8d59ca3361bcdaf4db729eeb4a71f714bd9feda)

s3:winbindd: implement wbint_bh_get_binding() in winbindd_dual_ndr.c

We use NCACN_INTERNAL here too...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 8c2aa56c7003e690252dfb05f78fd083c0c13dab)

s4:lib/messaging: implement irpc_bh_get_binding()

We just use NCACN_INTERNAL here...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit a6034592faf0676532c97a79b9c542bcd2de9537)

libcli/tstream_binding_handle: implement get_binding()

We just create an dummy binding handle here
as it's not really dcerpc...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 66411b96b88fbb7b405efa87015e4af1f2988fad)

librpc/rpc: add get_binding() to dcerpc_binding_handle_ops

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 1adf226e74375b2fb0932380f13bf19d8d856ea4)

s3:rpc_client: header signing is negotiated per transport connection

All gensec backends support GENSEC_FEATURE_SIGN_PKT_HEADER, so there's
no point in negotiating header signing based on the
auth context used during the DCERPC Bind.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 72b79a771fdbf54aa73b4c1b0372b9dffd81bfa6)

s3:rpc_client: make use of struct samba_sockaddr in rpc_pipe_open_ncalrpc()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit d406f9fffede27220f4d589c8ab3c7210ca01b74)

s3:winbindd: cm_connect_lsa_tcp() doesn't need to check for NCACN_IP_TCP or LEVEL_INTEGRITY

We only ever create domain->lsa_pipe_tcp with
cli_rpc_pipe_open_schannel_with_creds() and hardcoded NCACN_IP_TCP.

And schannel_update_internal returns an error with
an auth level lower than DCERPC_AUTH_LEVEL_INTEGRITY.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 61d38f70be378c59b42a6e20e7d49fea21255c6b)

s3:rpcclient: make use of dcerpc_binding_handle_auth_info()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 8dcbbebe56d3dad3650eca79322150e9d397e227)

s3:rpc_client: the transport_session_key is per connection!

It's not per auth_context_id, currently there's no difference
but that will change in future...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 1a311df3d61d3f3dbb986cc35470df6445802ff6)

s3:rpc_client: remove unused cli_get_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 3a3338bd184e3c51eab48fb1a03b1b3e749f2dc6)

s3:rpcclient/cli_drsuapi: make use of dcerpc_binding_handle_auth_session_key()

Note we only need to call this once per connection, not after
each request...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit f75189dee9db486d393cd17c25f5a84ce2d0b889)

s3:libnet_dssync: make use of dcerpc_binding_handle_auth_session_key()

Note we only need to call this once per connection, not after
each request...

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 099c8d904ece0edbefecbaa51fa50256fd2ea7c0)

s3:rpc_client: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 77da0914d5cccd7116ca3b6f75cb4d0cd0107239)

s3:utils/net_rpc: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit a14ac349deca11cda4159266cb3119fb3024999c)

s3:libnet_join: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 56175a0947fbb56d151fca057cf9696d55ea5d82)

s3:lib/netapi: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 2cb726e92c819d60509419a33cd8da265e1c99e7)

s4:torture/drs: make use of dcerpc_binding_handle_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit bbe529a2f6090d2b9df0f361aee9b0a28523cefe)

s4:pyrpc: make use of dcerpc_binding_handle_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 38874d28404f49a075bab71fb7357d266173791a)

s4:py_net: make use of dcerpc_binding_handle_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 5d3faa86fd47d6bc887e182f3eb5a51144339549)

s4:libnet: make use of dcerpc_binding_handle_auth_session_key() in libnet_become_dc.c

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit bf888b59e34169e1f6b9fd1d211cb8ff973cf333)

s4:drepl: make use of dcerpc_binding_handle_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 51f5eefcd119b5981281924a45380d99f9158a88)

s3:rpc_client: add rpccli_bh_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit af16ecece870846323956c37d3e15abe24e336fd)

s4:librpc/rpc: add dcerpc_bh_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 64467157479c04cb4bafa6c4ec19213009f0643f)

librpc/rpc: add dcerpc_binding_handle_auth_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 5e34173041eb04e30f3f6f81a3269d621220b09c)

s4:librpc/rpc: remove unused dcerpc_fetch_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit d9bf23f971dd8f426424f05076e76be587d42087)

s4:torture/rpc: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 4ccd2dd8a2f78bcd53c047889b8043e8d95a6d55)

xss4:pyrpc: make use of dcerpc_binding_handle_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 57905603e066e2d1ca60268b1b1e8166868d60e5)

s4:libnet: make use of dcerpc_binding_handle_transport_session_key() in libnet_passwd.c

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 93e8d58736e53d8d360ef2ee40f1619bd917c054)

s4:libnet: add struct dcerpc_binding_handle helper variables in libnet_passwd.c

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 2d1c182bf512666606b8192b2021d571a47d2f58)

s4:librpc/rpc: remove unused dcerpc_transport_encrypted()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 3cc6ddeb589f0357c436d1caa70130f053ddcba0)

s4:pyrpc: let py_iface_transport_encrypted() use dcerpc_binding_handle_transport_encrypted()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 9567d753ac0fcffb50f3710f3329334afb7595cd)

s3:rpc_client: add rpccli_bh_transport_session_key()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 5bb35cc7cca101390e58256e9b38e31113211e4f)

s4:librpc/rpc: add dcerpc_bh_transport_{encrypted,session_key}()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 9b50ab08c16c197291835c8e66cd064ca0bb6579)

librpc/rpc: add dcerpc_binding_handle_transport_{encrypted,session_key}()

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 3dc3f9cf1b6fc47baea329863e31ff140cf89550)

s4:torture/rpc: avoid using DCERPC_NDR_REF_ALLOC in fsrvp.c

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 061267e0e4de5c6379205a09376ec21374a87a3d)

s4:torture/rpc: remove useless usage of DCERPC_SIGN, DCERPC_SEAL

We already used DCERPC_AUTH_LEVEL_PRIVACY for the connection.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 136ee756fe3a41f863c2d1d538ad029bfe54e1d1)

s4:torture/rpc: avoid checking p->last_fault_code in iremotewinspool*

Now that we check for NT_STATUS_RPC_UNSUPPORTED_TYPE, there's no
need to check for DCERPC_NCA_S_UNSUPPORTED_TYPE.

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit b0be3c8af8240cfb7bc0b63fb7bf73f0f1655e4e)

librpc/rpc: map DCERPC_NCA_S_UNSUPPORTED_TYPE to NT_STATUS_RPC_UNSUPPORTED_TYPE

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 492f6a0eef7e7c64806ac9137c18a76fd116766d)

librpc/rpc: map DCERPC_NCA_S_SERVER_TOO_BUSY to NT_STATUS_RPC_SERVER_TOO_BUSY

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 722e93e4d9ace010e44cc041e6614ce5b71265d8)

VERSION: Bump version up to Samba 4.21.7...

and re-enable GIT_SNAPSHOT.

Signed-off-by: Jule Anger <janger@samba.org>

VERSION: Disable GIT_SNAPSHOT for the 4.21.6 release.

Signed-off-by: Jule Anger <janger@samba.org>

WHATSNEW: Add release notes for Samba 4.21.6.

Signed-off-by: Jule Anger <janger@samba.org>

CVE-2025-0620: smbd: smbd doesn't pick up group membership changes when re-authenticating an expired SMB session

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15707

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 6f0ae60428a024b4aba92a8103a698c1eca2357c)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Mon Jun  2 17:07:31 UTC 2025 on atb-devel-224

ctdb-daemon: Modernise some DEBUGs

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Thu May 29 10:57:35 UTC 2025 on atb-devel-224

(cherry picked from commit 5a582bddd834fffe2b27cc8b2e9468fa84dfc6f2)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Mon Jun  2 12:44:29 UTC 2025 on atb-devel-224

ctdb-daemon: Add configuration option shutdown extra timeout

See documentation change for details.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit 3a770c8d46934870f42059640b0aaa0c76a3f4fb)

ctdb-daemon: Run "startipreallocate" event in SHUTDOWN runstate

Even though all nodes may be shutting down there is still a very small
window for a race when multiple nodes are shut down.  For simplicity,
assume 2 nodes.  Assume the shutdowns of nodes are staggered, which is
usual because they're usually initiated by a loop (e.g. onnode -p all
ctdb shutdown).  Although commands can continue in parallel, some
commands are started later than others.

Consider this sequence:

1.  Node 0 reaches ctdb_shutdown_takeover() in
    ctdb_shutdown_sequence() and a takeover run starts

2.  Node 1 has not yet set its runlevel to SHUTDOWN in
    ctdb_shutdown_sequence()

3.  The leader node asks node 1 which IPs it can host

4.  Node 1 replies "all of them"

5.  Node 1 now sets its runlevel to SHUTDOWN in
    ctdb_shutdown_sequence()

6.  The leader node continues with the takeover run, first asking all
    nodes to run "startipreallocate"

7.  Node 0 runs "startipreallocate", so its NFS server starts grace

8.  Node 1 does not run "startipreallocate" because it is not in
    RUNNING runstate, so its NFS server does not start grace

9.  The leader node continues with the takeover run, first asking all
    nodes to run "releaseip" for IPs they can no longer hold

10. Node 0 releases all IPs, since it is SHUTDOWN runstate (so can't
    host IPs)

11. As part of this, the NFS server on node 0 releases locks held
    against IPs it is releasing

12. A client connected to node 1, where the NFS server is not in
    grace, takes ("steals") one of those locks

This client is then permitted to reclaim the lock when nodes are
restarted.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit 4877541cfd8f782f516f6471edc52629720963fb)

ctdb-daemon: Add configuration option shutdown failover timeout

Allows the timeout for failover during shutdown to be modified.
Defaults to 10s.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
SQ

Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit dd9b73119afd3a0c60c87c938b5aefc766ca78d2)

ctdb-daemon: Add failover on shutdown

Without this, NFS servers on other nodes will not go into grace before
this node releases locks.  This should also support improved behaviour
for SMB durable file handles.

The timeout is currently a constant 10s.  However, it will
subsequently be switched to an option.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit b84fbd7b3fedc998633400981ce0c5dc963d052e)

ctdb-protocol: Add CTDB server SRVID range

Normally, communication from other components to ctdbd is done via
controls.  However, there are contexts where receiving SRVID messages
in ctdbd makes sense, such as replies to outgoing SRVID messages.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit 631d1d38ad10c73aa559561bea6b5ed45c2226c4)

ctdb-daemon: Avoid aborting during early shutdown

An early shutdown can put ctdbd into SHUTDOWN runstate before ctdbd
has completed all early initialisation.  Some of the start-time
transitions then attempt to set the runstate to FIRST_RECOVERY or
RUNNING, which would make the runstate go backwards, so ctdbd aborts.

Upcoming changes cause ctdbd shutdown to take longer, so the problem
will become more likely.  With those changes, this can be
unreliably (50% of the time?)  triggered by:

  ctdb/tests/INTEGRATION/simple/cluster.091.version_check.sh

since it does an early shutdown due to a version mismatch.

Avoid this by noticing when the runstate is SHUTDOWN and refusing to
continue with subsequent early initialisation steps, which aren't
needed when shutting down.

Earlier runstate transitions do not seems likely to cause an abort
during early shutdown.  The following:

  ./tests/local_daemons.sh foo start 0; ./tests/local_daemons.sh foo stop 0

sees ctdbd already into FIRST_RECOVERY before the shutdown is
processed.

The change to ctdb_run_startup() probably isn't strictly necessary.
There will be no abort in this case.  ctdb_shutdown_sequence() will
always run the "shutdown" event and then stop the event daemon, so it
doesn't seem possible that services could be left running.  However,
we might as well avoid running the "startup" event when shutting down,
even if only to avoid confusing logs.

Ultimately, it seems like some redesign would be needed to avoid this
in a more predictable manner, rather than responding when an early
initialisation step inconveniently completes during shutdown.  For
example, hanging a lot of the start-time event handling off a common
talloc context, could allow it to be cancelled with a single
TALLOC_FREE().  However, a change like that would involve a lot of
analysis to ensure that the talloc hierarchy is correct and there is
no change of free'd pointers being dereferenced.  So, we're probably
better off just keeping this issue in mind during a broader redesign.

This workaround appears to be sufficient.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15858

Signed-off-by: Martin Schwenke <mschwenke@ddn.com>
Reviewed-by: Amitay Isaacs <amitay@gmail.com>
(cherry picked from commit c03e6b9d50cac67fe33dc6b120996d1915331be6)

vfs_ceph_snapshots: Always calculate absolute snapshot path

Use the same logic from shadow_copy2 module to always prepend the
connectpath to the relative snapshot path so as to return converted
path corresponding to the file's share root.

Please note that with the current working directory staying at the
connectpath level we are safe to prefix it to the smb_filename. In
other words it seems we never get past the connectpath internally
during normal file system operations via chdir(). Since all relative
paths are now based on dirfsp we could constitute absolute path by
prepending the connectpath to full_path_from_dirfsp_atname() output
ignoring the current working directory.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15819

Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: David Disseldorp <ddiss@samba.org>
Autobuild-User(master): Anoop C S <anoopcs@samba.org>
Autobuild-Date(master): Wed Apr 30 11:32:59 UTC 2025 on atb-devel-224

(cherry picked from commit 95a2b50b1983a6ba810a96f50b27db7c992c02c0)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Thu May  8 12:46:14 UTC 2025 on atb-devel-224

vfs_ceph_snapshots: Use full path from dirfsp at smb_fname

In ceph_snap_gmt_openat() we hand in the incoming smb_fname as it is
to ceph_snap_gmt_strip_snapshot() which is then passed on to derive
the actual snapshot path using ceph_snap_gmt_convert(). But this can
go wrong in ceph_snap_gmt_convert_dir() while opening the snapdir.
Unless we constitute the full path from dirfsp at the first place we
always end up opening the snapdir from the parent directory with
OpenDir().

For example with dirfsp("foobar") and smb_fname("shift.txt"), we open
snapdir from share root because parent is calculated as empty string
via ceph_snap_get_parent_path(). Instead we could construct the full
path from dirfsp using full_path_from_dirfsp_atname() to ensure we
don't open the wrong snapdir.

Since we have access to the twrp token at VFS layer it doesn't make
much sense to make use of ceph_snap_gmt_strip_snapshot() in openat.
We could instead directly act based on already available twrp token
avoiding an extra copy of incoming smb_filename.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15819

Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: David Disseldorp ddiss@samba.org
(cherry picked from commit ad8b2dbb67d87db22f2fa2df814dd91cbe071e60)

lib/krb5_wrap: Fix placement of TALLOC_FREE(frame)

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15727

Signed-off-by: Pavel Filipenský <pfilipensky@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Apr 23 12:14:17 UTC 2025 on atb-devel-224

(cherry picked from commit b5bd36dfd7dfd9a09a3f9378330df3be9db4551f)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Wed Apr 30 16:42:48 UTC 2025 on atb-devel-224

s3:rpc_server: make sure we can bind to the same port on all ip addresses

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15851

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 783ca9dc434bd1d18b762185ee936fcbcf292124)

vfs: Fix "wide links = yes"

vfs_wide_links hides symlinks from the rest of smbd, and it implicitly
follows symlinks. Also, O_PATH will expose symlinks to the rest of
smbd, remove that.

We also need to do this for posix paths, as deep inside
rename_internals we want to avoid case-insensitive lookups by setting
SMB_FILENAME_POSIX_PATH.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=15841

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Apr 16 20:56:33 UTC 2025 on atb-devel-224

(cherry picked from commit 9e637a28bd0b5adc2d90df9ca4a1c864a648b0f4)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Thu Apr 17 15:00:12 UTC 2025 on atb-devel-224

python:gp_cert_auto_enrol: fix GUID stringification

We were using some broken ad-hoc unpacking to do what the ndr
unpacker does perfectly well.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15839

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
Autobuild-User(master): Douglas Bagnall <dbagnall@samba.org>
Autobuild-Date(master): Tue Mar 25 05:21:49 UTC 2025 on atb-devel-224

(cherry picked from commit 47ff42232048c008a7b361a948e5ac79311b5458)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Thu Apr 17 12:53:59 UTC 2025 on atb-devel-224

samba-tool gpo: better entities check copes with new lines

Per https://www.w3.org/TR/xml/#sec-entity-decl (and MS references)
there is always some whitespace between '<!ENTITY' and the name, and
between the name and whatever is next. Also, it is valid XML to have
newlines inside entity declarations, like this:

<!ENTITY
    bubble
      "*S-1-5-113"
      >

We used to create such files, so we should allow them.

There is a kind of entity that has '%' before the name, and there are
non-ascii names, which we continue not to support.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15829

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
(cherry picked from commit 6107656ebc8d092b2c1907940b2486ab0265aad9)

samba-tool gpo backup fix --generalize

This was broken with commit ce56d336f234febfd4cb3da11dd584842c24ce1d
but we didn't notice because the test was already broken.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15829

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
(cherry picked from commit 65751f2562f98bd7fd0734dc00784e6395d76322)

pytest: samba-tool gpo: fix has_difference(sortlines=True)

We had

             file1 = open(path1).readlines()
             file1.sort()
             file2 = open(path1).readlines()
             file2.sort()

which is opening path1 in both cases.

This meant we were testing nothing because the assertions are all that
the files are the same -- though the only affected check is one in
test_backup_restore_generalize().

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15829

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
(cherry picked from commit 6b619b568f6661d3a5f0701cdfaf1e1e4943ff6f)

python:netcmd:gpo: fix crash when updating an MOTD GPO

When the policy exists already, there is no exception and the code
tries to use the "data" variable, but it doesn't exist because it was
only defined in the exception handling.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15774

Signed-off-by: Andreas Hasenack <andreas.hasenack@canonical.com>
Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
(cherry picked from commit e87e20c04d90292e3a5caac8ea3105b16f948ed3)

pytest: check we can set GPO more than once

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15774

Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Reviewed-by: Jennifer Sutton <jennifersutton@catalyst.net.nz>
(cherry picked from commit 969cb41e06247949c3992cab25e824795204e31e)

s3:rpc_server/srvsvc: use brl_get_locks_readonly() instead of brl_get_locks()

No need to keep the record locked longer then needed.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit c36cc2b6720a2cfe54ce52a500dc499418e27e34)

smbd: use share_mode_do_locked_brl() in vfs_default_durable_reconnect()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(backported from commit dc03a06ffcc79d0818ae4a36fe3f2df705144138)
[slow@samba.org: conflict due to removed delayed write time handling]
[slow@samba.org: conflict due to filename_convert_dirfsp_rel()]

smbd: use share_mode_do_locked_brl() in vfs_default_durable_disconnect()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(backported from commit 393379fc9c726eb781fd1bfb3a70ea2802739aff)
[slow@samba.org: conflict due to removed delayed write time handling]

smbd: use share_mode_do_locked_brl() in strict_lock_check_default()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 56bb20c87a733ab8f7efedd881ea0ecaf51b2ba8)

smbd: check can_lock in strict_lock_check_default()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 678f28c1af7c160ffdcb0e4baa0a7d4b9906f2e5)

s3/locking: prepare brl_locktest() for upgradable read-only locks

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 8f9387ceb5c94c7db92ab342e33c64b858c301b1)

smbd: call locking_close_file() while still holding a glock on the locking.tdb record

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 4d680b6c17ee7674b9686aec2b69038f89e1989a)

s3/brlock: remove brl_get_locks_for_locking()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Pair-Programmed-With: Ralph Boehme <slow@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 0c4c430c50e15d591a0d871a5f3e59e8be0d0a83)

smbd: use share_mode_do_locked_brl()

Fix a deadlock that can happen if two clients happen to open and byte-range-lock
two different files whos record in locking.tdb and brlock.tdb happen to sit on
the same hashchain.

The deadlock was introduced by commit
680c7907325b433856ac1dd916ab63e671fbe4ab. Before, we used share_mode_do_locked()
in do_lock() which meant we acquired a chainlock on locking.tdb before getting a
chainlock on brlock.tdb via brl_get_locks_for_locking(), so the TDB chainlock
order invariant was always uphold.

The following race between specific client requests lead to the deadlock.

Client A) issues a byte-range-lock request on a file:

A1) glock locking.tdb (via _share_mode_do_locked_vfs_allowed())
A2) chainlock brlock.tdb (via brl_lock())
A3) attempt to chainlock locking.tdb (via share_mode_g_lock_dump())
[1]

Client B) opens a different (!) file:

B1) glock and chainlock locking.tdb (via _share_mode_entry_prepare_lock())
B2) attempt to chainlock brlock.tdb (via file_has_brlocks())
[2]

The glock from A1 is per record and hence doesn't synchronize with the glock
from B1 as it is for a different file and hence a different record, subsequently
A2 and A3 violate the lock order constraint

To avoid the chainlock lock order violation in the second client we modify the
br-lock code to not take the brlock.tdb chainlock from step A2 via
br_get_locks() for the whole time we process the request. Instead we just fetch
the br-locks via br_get_locks_readonly(), so when running into
contend_level2_oplocks_begin_default() to check for leases and looking into
locking.tdb we don't hold a brlock.tdb chainlock.

Or im simpler terms, we only ever take at most one low-level TDB chainlock at a
time:

Byte-range-lock code calls share_mode_do_locked_brl(..., cb_fn, ...):
1) chainlock locking.tdb
2) glock locking.tdb (via share_mode_do_locked_vfs_allowed())
3) chainunlock locking.tdb
4) share_mode_do_locked_brl_fn() -> brl_get_locks_readonly_parse():
   a) chainlock brlock.tdb
   b) parse record and store in-memory copy
   c) chainunlock brlock.tdb
5) run cb_fn()
6) chainlock brlock.tdb:
   a) br_lck->record = dbwrap_fetch_locked(brlock_db, ...)
   b) store modifed br_lck from 5) via byte_range_lock_flush()
7) chainunlock brlock.tdb
8) chainlock locking.tdb
9) gunlock locking.tdb
10) chainunlock locking.tdb

All access to brlock.tdb is synchronized correctly via glocks on the locking.tdb
record of the file (step 3)), so operations still appear atomic to clients.

As a result of using share_mode_do_locked_brl(), the functions do_[un]lock() ->
brl_[un]lock() now loop over the same br_lck object in memory, avoiding
repeatedly fetching and storing the locks per loop.

[1]
Full SBT:

  #0  0x00007fffa0cecbb0 in __pthread_mutex_lock_full () from /lib64/glibc-hwcaps/power9/libpthread-2.28.so
  #1  0x00007fffa0a73cf8 in chain_mutex_lock (m=<optimized out>, m@entry=0x7fff9ae071b0, waitflag=<optimized out>, waitflag@entry=true) at ../../lib/tdb/common/mutex.c:182
  #2  0x00007fffa0a7432c in tdb_mutex_lock (tdb=0x1543ba120, rw=<optimized out>, off=<optimized out>, len=<optimized out>, waitflag=<optimized out>, pret=0x7fffd7df3858) at ../../lib/tdb/common/mutex.c:234
  #3  0x00007fffa0a6812c in fcntl_lock (waitflag=<optimized out>, len=1, off=376608, rw=0, tdb=0x1543ba120) at ../../lib/tdb/common/lock.c:200
  #4  tdb_brlock (tdb=0x1543ba120, rw_type=<optimized out>, offset=<optimized out>, len=1, flags=<optimized out>) at ../../lib/tdb/common/lock.c:200
  #5  0x00007fffa0a68af8 in tdb_nest_lock (flags=<optimized out>, ltype=0, offset=<optimized out>, tdb=0x1543ba120) at ../../lib/tdb/common/lock.c:390
  #6  tdb_nest_lock (tdb=0x1543ba120, offset=<optimized out>, ltype=<optimized out>, flags=<optimized out>) at ../../lib/tdb/common/lock.c:336
  #7  0x00007fffa0a69088 in tdb_lock_list (tdb=0x1543ba120, list=<optimized out>, ltype=<optimized out>, waitflag=<optimized out>) at ../../lib/tdb/common/lock.c:482
  #8  0x00007fffa0a69198 in tdb_lock (tdb=0x1543ba120, list=<optimized out>, ltype=<optimized out>) at ../../lib/tdb/common/lock.c:500
  #9  0x00007fffa0a64b50 in tdb_find_lock_hash (tdb=<optimized out>, tdb@entry=0x1543ba120, key=..., hash=<optimized out>, locktype=<optimized out>, locktype@entry=0, rec=<optimized out>, rec@entry=0x7fffd7df3ab0) at ../../lib/tdb/common/tdb.c:165
  #10 0x00007fffa0a64ed0 in tdb_parse_record (tdb=0x1543ba120, key=..., parser=0x7fffa0e74470 <db_ctdb_ltdb_parser>, private_data=0x7fffd7df3b18) at ../../lib/tdb/common/tdb.c:329
  #11 0x00007fffa0e74cbc in db_ctdb_ltdb_parse (db=<optimized out>, private_data=0x7fffd7df3b70, parser=0x7fffa0e76470 <db_ctdb_parse_record_parser_nonpersistent>, key=...) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:170
  #12 db_ctdb_try_parse_local_record (ctx=ctx@entry=0x1543d4580, key=..., state=state@entry=0x7fffd7df3b70) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:1385
  #13 0x00007fffa0e76024 in db_ctdb_parse_record (db=<optimized out>, key=..., parser=0x7fffa1313910 <dbwrap_watched_parse_record_parser>, private_data=0x7fffd7df3c08) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:1425
  #14 0x00007fffa0884760 in dbwrap_parse_record (db=<optimized out>, key=..., parser=<optimized out>, private_data=<optimized out>) at ../../lib/dbwrap/dbwrap.c:454
  #15 0x00007fffa1313ab4 in dbwrap_watched_parse_record (db=0x1543a7160, key=..., parser=0x7fffa13187d0 <g_lock_dump_fn>, private_data=0x7fffd7df3ce8) at ../../source3/lib/dbwrap/dbwrap_watch.c:783
  #16 0x00007fffa0884760 in dbwrap_parse_record (db=<optimized out>, key=..., parser=<optimized out>, private_data=<optimized out>) at ../../lib/dbwrap/dbwrap.c:454
  #17 0x00007fffa131c004 in g_lock_dump (ctx=<error reading variable: value has been optimized out>, key=..., fn=0x7fffa14f3d70 <fsp_update_share_mode_flags_fn>, private_data=0x7fffd7df3dd8) at ../../source3/lib/g_lock.c:1653
  #18 0x00007fffa14f434c in share_mode_g_lock_dump (key=..., fn=0x7fffa14f3d70 <fsp_update_share_mode_flags_fn>, private_data=0x7fffd7df3dd8) at ../../source3/locking/share_mode_lock.c:96
  #19 0x00007fffa14f8d44 in fsp_update_share_mode_flags (fsp=0x15433c550) at ../../source3/locking/share_mode_lock.c:1181
  #20 file_has_read_lease (fsp=0x15433c550) at ../../source3/locking/share_mode_lock.c:1207
  #21 0x00007fffa15ccc98 in contend_level2_oplocks_begin_default (type=<optimized out>, fsp=0x15433c550) at ../../source3/smbd/smb2_oplock.c:1282
  #22 smbd_contend_level2_oplocks_begin (fsp=0x15433c550, type=<optimized out>) at ../../source3/smbd/smb2_oplock.c:1338
  #23 0x00007fffa0dd0b54 in contend_level2_oplocks_begin (fsp=<optimized out>, type=<optimized out>) at ../../source3/lib/smbd_shim.c:72
  #24 0x00007fffa14ecfd0 in brl_lock_windows_default (br_lck=0x154421330, plock=0x7fffd7df4250) at ../../source3/locking/brlock.c:457
  #25 0x00007fffa150b70c in vfswrap_brl_lock_windows (handle=<optimized out>, br_lck=<optimized out>, plock=<optimized out>) at ../../source3/modules/vfs_default.c:3424
  #26 0x00007fffa1561910 in smb_vfs_call_brl_lock_windows (handle=<optimized out>, br_lck=<optimized out>, plock=<optimized out>) at ../../source3/smbd/vfs.c:2686
  #27 0x00007fff9c0a7350 in smb_time_audit_brl_lock_windows (handle=<optimized out>, br_lck=0x154421330, plock=0x7fffd7df4250) at ../../source3/modules/vfs_time_audit.c:1740
  #28 0x00007fffa1561910 in smb_vfs_call_brl_lock_windows (handle=<optimized out>, br_lck=<optimized out>, plock=<optimized out>) at ../../source3/smbd/vfs.c:2686
  #29 0x00007fffa14ed410 in brl_lock (br_lck=0x154421330, smblctx=3102281601, pid=..., start=0, size=18446744073709551615, lock_type=<optimized out>, lock_flav=WINDOWS_LOCK, blocker_pid=0x7fffd7df4540, psmblctx=0x7fffd7df4558) at ../../source3/locking/brlock.c:1004
  #30 0x00007fffa14e7b18 in do_lock_fn (lck=<optimized out>, private_data=0x7fffd7df4508) at ../../source3/locking/locking.c:271
  #31 0x00007fffa14fcd94 in _share_mode_do_locked_vfs_allowed (id=..., fn=0x7fffa14e7a60 <do_lock_fn>, private_data=0x7fffd7df4508, location=<optimized out>) at ../../source3/locking/share_mode_lock.c:2927
  #32 0x00007fffa14e918c in do_lock (fsp=0x15433c550, req_mem_ctx=<optimized out>, req_guid=<optimized out>, smblctx=<optimized out>, count=18446744073709551615, offset=0, lock_type=<optimized out>, lock_flav=<optimized out>, pblocker_pid=0x7fffd7df46f0,
      psmblctx=0x7fffd7df46d8) at ../../source3/locking/locking.c:335
  #33 0x00007fffa155381c in smbd_do_locks_try (fsp=0x15433c550, num_locks=<optimized out>, locks=0x1543bc310, blocker_idx=0x7fffd7df46d6, blocking_pid=0x7fffd7df46f0, blocking_smblctx=0x7fffd7df46d8) at ../../source3/smbd/blocking.c:46
  #34 0x00007fffa159dc90 in smbd_smb2_lock_try (req=req@entry=0x1543bc080) at ../../source3/smbd/smb2_lock.c:590
  #35 0x00007fffa159ee8c in smbd_smb2_lock_send (in_locks=<optimized out>, in_lock_count=1, in_lock_sequence=<optimized out>, fsp=0x15433c550, smb2req=0x1543532e0, ev=0x154328120, mem_ctx=0x1543532e0) at ../../source3/smbd/smb2_lock.c:488
  #36 smbd_smb2_request_process_lock (req=0x1543532e0) at ../../source3/smbd/smb2_lock.c:150
  #37 0x00007fffa158a368 in smbd_smb2_request_dispatch (req=0x1543532e0) at ../../source3/smbd/smb2_server.c:3515
  #38 0x00007fffa158c540 in smbd_smb2_io_handler (fde_flags=<optimized out>, xconn=0x154313f30) at ../../source3/smbd/smb2_server.c:5112
  #39 smbd_smb2_connection_handler (ev=<optimized out>, fde=<optimized out>, flags=<optimized out>, private_data=<optimized out>) at ../../source3/smbd/smb2_server.c:5150
  #40 0x00007fffa1198b2c in tevent_common_invoke_fd_handler (fde=0x1543670f0, flags=<optimized out>, removed=0x0) at ../../lib/tevent/tevent_fd.c:158
  #41 0x00007fffa11a2b9c in epoll_event_loop (tvalp=0x7fffd7df4b28, epoll_ev=0x1543b4e80) at ../../lib/tevent/tevent_epoll.c:730
  #42 epoll_event_loop_once (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent_epoll.c:946
  #43 0x00007fffa11a0090 in std_event_loop_once (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent_standard.c:110
  #44 0x00007fffa119744c in _tevent_loop_once (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent.c:823
  #45 0x00007fffa1197884 in tevent_common_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:950
  #46 0x00007fffa119ffc0 in std_event_loop_wait (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent_standard.c:141
  #47 0x00007fffa1197978 in _tevent_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:971
  #48 0x00007fffa15737fc in smbd_process (ev_ctx=0x154328120, msg_ctx=<optimized out>, sock_fd=<optimized out>, interactive=<optimized out>) at ../../source3/smbd/smb2_process.c:2158
  #49 0x000000011db5c554 in smbd_accept_connection (ev=0x154328120, fde=<optimized out>, flags=<optimized out>, private_data=<optimized out>) at ../../source3/smbd/server.c:1150
  #50 0x00007fffa1198b2c in tevent_common_invoke_fd_handler (fde=0x1543ac2d0, flags=<optimized out>, removed=0x0) at ../../lib/tevent/tevent_fd.c:158
  #51 0x00007fffa11a2b9c in epoll_event_loop (tvalp=0x7fffd7df4f98, epoll_ev=0x154328350) at ../../lib/tevent/tevent_epoll.c:730
  #52 epoll_event_loop_once (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent_epoll.c:946
  #53 0x00007fffa11a0090 in std_event_loop_once (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent_standard.c:110
  #54 0x00007fffa119744c in _tevent_loop_once (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent.c:823
  #55 0x00007fffa1197884 in tevent_common_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:950
  #56 0x00007fffa119ffc0 in std_event_loop_wait (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent_standard.c:141
  #57 0x00007fffa1197978 in _tevent_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:971
  #58 0x000000011db58c54 in smbd_parent_loop (parent=<optimized out>, ev_ctx=0x154328120) at ../../source3/smbd/server.c:1499
  #59 main (argc=<optimized out>, argv=<optimized out>) at ../../source3/smbd/server.c:2258

[2]
Full SBT:

  #0  0x00007fffa0cecbb0 in __pthread_mutex_lock_full () from /lib64/glibc-hwcaps/power9/libpthread-2.28.so
  #1  0x00007fffa0a73cf8 in chain_mutex_lock (m=<optimized out>, m@entry=0x7fff9b3a71b0, waitflag=<optimized out>, waitflag@entry=true) at ../../lib/tdb/common/mutex.c:182
  #2  0x00007fffa0a7432c in tdb_mutex_lock (tdb=0x1543c6900, rw=<optimized out>, off=<optimized out>, len=<optimized out>, waitflag=<optimized out>, pret=0x7fffd7df2e28) at ../../lib/tdb/common/mutex.c:234
  #3  0x00007fffa0a6812c in fcntl_lock (waitflag=<optimized out>, len=1, off=376608, rw=0, tdb=0x1543c6900) at ../../lib/tdb/common/lock.c:200
  #4  tdb_brlock (tdb=0x1543c6900, rw_type=<optimized out>, offset=<optimized out>, len=1, flags=<optimized out>) at ../../lib/tdb/common/lock.c:200
  #5  0x00007fffa0a68af8 in tdb_nest_lock (flags=<optimized out>, ltype=0, offset=<optimized out>, tdb=0x1543c6900) at ../../lib/tdb/common/lock.c:390
  #6  tdb_nest_lock (tdb=0x1543c6900, offset=<optimized out>, ltype=<optimized out>, flags=<optimized out>) at ../../lib/tdb/common/lock.c:336
  #7  0x00007fffa0a69088 in tdb_lock_list (tdb=0x1543c6900, list=<optimized out>, ltype=<optimized out>, waitflag=<optimized out>) at ../../lib/tdb/common/lock.c:482
  #8  0x00007fffa0a69198 in tdb_lock (tdb=0x1543c6900, list=<optimized out>, ltype=<optimized out>) at ../../lib/tdb/common/lock.c:500
  #9  0x00007fffa0a64b50 in tdb_find_lock_hash (tdb=<optimized out>, tdb@entry=0x1543c6900, key=..., hash=<optimized out>, locktype=<optimized out>, locktype@entry=0, rec=<optimized out>, rec@entry=0x7fffd7df3080) at ../../lib/tdb/common/tdb.c:165
  #10 0x00007fffa0a64ed0 in tdb_parse_record (tdb=0x1543c6900, key=..., parser=0x7fffa0e74470 <db_ctdb_ltdb_parser>, private_data=0x7fffd7df30e8) at ../../lib/tdb/common/tdb.c:329
  #11 0x00007fffa0e74cbc in db_ctdb_ltdb_parse (db=<optimized out>, private_data=0x7fffd7df3140, parser=0x7fffa0e76470 <db_ctdb_parse_record_parser_nonpersistent>, key=...) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:170
  #12 db_ctdb_try_parse_local_record (ctx=ctx@entry=0x154328fc0, key=..., state=state@entry=0x7fffd7df3140) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:1385
  #13 0x00007fffa0e76024 in db_ctdb_parse_record (db=<optimized out>, key=..., parser=0x7fffa14ec820 <brl_get_locks_readonly_parser>, private_data=0x7fffd7df3218) at ../../source3/lib/dbwrap/dbwrap_ctdb.c:1425
  #14 0x00007fffa0884760 in dbwrap_parse_record (db=<optimized out>, key=..., parser=<optimized out>, private_data=<optimized out>) at ../../lib/dbwrap/dbwrap.c:454
  #15 0x00007fffa14ef5bc in brl_get_locks_readonly (fsp=0x1543d01e0) at ../../source3/locking/brlock.c:1884
  #16 0x00007fffa1546968 in file_has_brlocks (fsp=0x1543d01e0) at ../../source3/smbd/open.c:2232
  #17 delay_for_oplock (pgranted=<synthetic pointer>, poplock_type=<synthetic pointer>, first_open_attempt=<optimized out>, create_disposition=1, have_sharing_violation=false, lck=0x7fffd7df3ce8, lease=0x0, oplock_request=0, fsp=0x1543d01e0) at ../../source3/smbd/open.c:2749
  #18 handle_share_mode_lease (pgranted=<synthetic pointer>, poplock_type=<synthetic pointer>, first_open_attempt=<optimized out>, lease=0x0, oplock_request=0, share_access=7, access_mask=131201, create_disposition=1, lck=0x7fffd7df3ce8, fsp=0x1543d01e0) at ../../source3/smbd/open.c:2865
  #19 check_and_store_share_mode (first_open_attempt=<optimized out>, lease=0x0, oplock_request=0, share_access=7, access_mask=131201, create_disposition=1, lck=0x7fffd7df3ce8, req=0x154414800, fsp=0x1543d01e0) at ../../source3/smbd/open.c:3333
  #20 open_ntcreate_lock_add_entry (lck=0x7fffd7df3ce8, keep_locked=0x7fffd7df3ad0, private_data=0x7fffd7df3cc8) at ../../source3/smbd/open.c:3688
  #21 0x00007fffa14f6248 in share_mode_entry_prepare_lock_fn (glck=0x7fffd7df35b8, cb_private=0x7fffd7df3a88) at ../../source3/locking/share_mode_lock.c:2978
  #22 0x00007fffa1317680 in g_lock_lock_cb_run_and_store (cb_state=cb_state@entry=0x7fffd7df35b8) at ../../source3/lib/g_lock.c:597
  #23 0x00007fffa1319df8 in g_lock_lock_simple_fn (rec=0x7fffd7df3798, value=..., private_data=0x7fffd7df39a0) at ../../source3/lib/g_lock.c:1212
  #24 0x00007fffa13160e0 in dbwrap_watched_do_locked_fn (backend_rec=<optimized out>, backend_value=..., private_data=0x7fffd7df3768) at ../../source3/lib/dbwrap/dbwrap_watch.c:458
  #25 0x00007fffa0884e48 in dbwrap_do_locked (db=<optimized out>, key=..., fn=0x7fffa1316080 <dbwrap_watched_do_locked_fn>, private_data=0x7fffd7df3768) at ../../lib/dbwrap/dbwrap.c:602
  #26 0x00007fffa1315274 in dbwrap_watched_do_locked (db=0x1543a7160, key=..., fn=0x7fffa1319ca0 <g_lock_lock_simple_fn>, private_data=0x7fffd7df39a0) at ../../source3/lib/dbwrap/dbwrap_watch.c:480
  #27 0x00007fffa0884d60 in dbwrap_do_locked (db=<optimized out>, key=..., fn=<optimized out>, private_data=<optimized out>) at ../../lib/dbwrap/dbwrap.c:582
  #28 0x00007fffa131b458 in g_lock_lock (ctx=0x1543cc630, key=..., type=<optimized out>, timeout=..., cb_fn=0x7fffa14f6190 <share_mode_entry_prepare_lock_fn>, cb_private=0x7fffd7df3a88) at ../../source3/lib/g_lock.c:1267
  #29 0x00007fffa14fd060 in _share_mode_entry_prepare_lock (prepare_state=0x7fffd7df3cc8, id=..., servicepath=<optimized out>, smb_fname=<optimized out>, old_write_time=<optimized out>, fn=<optimized out>, private_data=0x7fffd7df3cc8, location=0x7fffa165b880 "../../source3/smbd/open.c:4292") at ../../source3/locking/share_mode_lock.c:3033
  #30 0x00007fffa15491e0 in open_file_ntcreate (conn=conn@entry=0x154382050, req=req@entry=0x154414800, access_mask=<optimized out>, access_mask@entry=131201, share_access=share_access@entry=7, create_disposition=create_disposition@entry=1, create_options=create_options@entry=0, new_dos_attributes=<optimized out>, new_dos_attributes@entry=128, oplock_request=oplock_request@entry=0, lease=<optimized out>, lease@entry=0x0, private_flags=<optimized out>, private_flags@entry=0, parent_dir_fname=<optimized out>, smb_fname_atname=<optimized out>, pinfo=<optimized out>, pinfo@entry=0x7fffd7df3f1c, fsp=<optimized out>, fsp@entry=0x1543d01e0) at ../../source3/smbd/open.c:4286
  #31 0x00007fffa154b94c in create_file_unixpath (conn=conn@entry=0x154382050, req=req@entry=0x154414800, dirfsp=dirfsp@entry=0x15439a7f0, smb_fname=smb_fname@entry=0x154416300, access_mask=access_mask@entry=131201, share_access=share_access@entry=7, create_disposition=create_disposition@entry=1, create_options=create_options@entry=0, file_attributes=file_attributes@entry=128, oplock_request=<optimized out>, oplock_request@entry=0, lease=<optimized out>, lease@entry=0x0, allocation_size=allocation_size@entry=0, private_flags=private_flags@entry=0, sd=sd@entry=0x0, ea_list=ea_list@entry=0x0, result=result@entry=0x7fffd7df4168, pinfo=pinfo@entry=0x7fffd7df4160) at ../../source3/smbd/open.c:6290
  #32 0x00007fffa154dfac in create_file_default (conn=0x154382050, req=0x154414800, dirfsp=0x15439a7f0, smb_fname=0x154416300, access_mask=<optimized out>, share_access=<optimized out>, create_disposition=<optimized out>, create_options=<optimized out>, file_attributes=128, oplock_request=0, lease=0x0, allocation_size=0, private_flags=0, sd=0x0, ea_list=0x0, result=0x1544144e8, pinfo=0x1544144fc, in_context_blobs=0x7fffd7df4798, out_context_blobs=0x154414710) at ../../source3/smbd/open.c:6609
  #33 0x00007fffa150972c in vfswrap_create_file (handle=<optimized out>, req=<optimized out>, dirfsp=<optimized out>, smb_fname=<optimized out>, access_mask=<optimized out>, share_access=<optimized out>, create_disposition=<optimized out>, create_options=<optimized out>, file_attributes=128, oplock_request=0, lease=0x0, allocation_size=0, private_flags=0, sd=0x0, ea_list=0x0, result=0x1544144e8, pinfo=0x1544144fc, in_context_blobs=0x7fffd7df4798, out_context_blobs=0x154414710) at ../../source3/modules/vfs_default.c:776
  #34 0x00007fffa1559cbc in smb_vfs_call_create_file (handle=<optimized out>, req=<optimized out>, dirfsp=<optimized out>, smb_fname=<optimized out>, access_mask=<optimized out>, share_access=<optimized out>, create_disposition=<optimized out>, create_options=<optimized out>, file_attributes=128, oplock_request=0, lease=0x0, allocation_size=0, private_flags=0, sd=0x0, ea_list=0x0, result=0x1544144e8, pinfo=0x1544144fc, in_context_blobs=0x7fffd7df4798, out_context_blobs=0x154414710) at ../../source3/smbd/vfs.c:1560
  #35 0x00007fff9c0a9ec4 in smb_time_audit_create_file (handle=0x154426820, req=0x154414800, dirfsp=0x15439a7f0, fname=0x154416300, access_mask=<optimized out>, share_access=<optimized out>, create_disposition=<optimized out>, create_options=<optimized out>, file_attributes=128, oplock_request=0, lease=0x0, allocation_size=0, private_flags=0, sd=0x0, ea_list=0x0, result_fsp=0x1544144e8, pinfo=0x1544144fc, in_context_blobs=0x7fffd7df4798, out_context_blobs=0x154414710) at ../../source3/modules/vfs_time_audit.c:634
  #36 0x00007fffa1559cbc in smb_vfs_call_create_file (handle=<optimized out>, req=<optimized out>, dirfsp=<optimized out>, smb_fname=<optimized out>, access_mask=<optimized out>, share_access=<optimized out>, create_disposition=<optimized out>, create_options=<optimized out>, file_attributes=128, oplock_request=0, lease=0x0, allocation_size=0, private_flags=0, sd=0x0, ea_list=0x0, result=0x1544144e8, pinfo=0x1544144fc, in_context_blobs=0x7fffd7df4798, out_context_blobs=0x154414710) at ../../source3/smbd/vfs.c:1560
  #37 0x00007fffa1597aa8 in smbd_smb2_create_send (in_context_blobs=..., in_name=0x154413ca0, in_create_options=<optimized out>, in_create_disposition=<optimized out>, in_share_access=<optimized out>, in_file_attributes=<optimized out>, in_desired_access=<optimized out>, in_impersonation_level=<optimized out>, in_oplock_level=<optimized out>, smb2req=0x154413770, ev=0x154328120, mem_ctx=0x154413770) at ../../source3/smbd/smb2_create.c:1115
  #38 smbd_smb2_request_process_create (smb2req=0x154413770) at ../../source3/smbd/smb2_create.c:291
  #39 0x00007fffa158a628 in smbd_smb2_request_dispatch (req=0x154413770) at ../../source3/smbd/smb2_server.c:3485
  #40 0x00007fffa158c540 in smbd_smb2_io_handler (fde_flags=<optimized out>, xconn=0x154313f30) at ../../source3/smbd/smb2_server.c:5112
  #41 smbd_smb2_connection_handler (ev=<optimized out>, fde=<optimized out>, flags=<optimized out>, private_data=<optimized out>) at ../../source3/smbd/smb2_server.c:5150
  #42 0x00007fffa1198b2c in tevent_common_invoke_fd_handler (fde=0x15435add0, flags=<optimized out>, removed=0x0) at ../../lib/tevent/tevent_fd.c:158
  #43 0x00007fffa11a2b9c in epoll_event_loop (tvalp=0x7fffd7df4b28, epoll_ev=0x1543b4e80) at ../../lib/tevent/tevent_epoll.c:730
  #44 epoll_event_loop_once (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent_epoll.c:946
  #45 0x00007fffa11a0090 in std_event_loop_once (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent_standard.c:110
  #46 0x00007fffa119744c in _tevent_loop_once (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent.c:823
  #47 0x00007fffa1197884 in tevent_common_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:950
  #48 0x00007fffa119ffc0 in std_event_loop_wait (ev=0x154328120, location=0x7fffa1668db8 "../../source3/smbd/smb2_process.c:2158") at ../../lib/tevent/tevent_standard.c:141
  #49 0x00007fffa1197978 in _tevent_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:971
  #50 0x00007fffa15737fc in smbd_process (ev_ctx=0x154328120, msg_ctx=<optimized out>, sock_fd=<optimized out>, interactive=<optimized out>) at ../../source3/smbd/smb2_process.c:2158
  #51 0x000000011db5c554 in smbd_accept_connection (ev=0x154328120, fde=<optimized out>, flags=<optimized out>, private_data=<optimized out>) at ../../source3/smbd/server.c:1150
  #52 0x00007fffa1198b2c in tevent_common_invoke_fd_handler (fde=0x1543ac2d0, flags=<optimized out>, removed=0x0) at ../../lib/tevent/tevent_fd.c:158
  #53 0x00007fffa11a2b9c in epoll_event_loop (tvalp=0x7fffd7df4f98, epoll_ev=0x154328350) at ../../lib/tevent/tevent_epoll.c:730
  #54 epoll_event_loop_once (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent_epoll.c:946
  #55 0x00007fffa11a0090 in std_event_loop_once (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent_standard.c:110
  #56 0x00007fffa119744c in _tevent_loop_once (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent.c:823
  #57 0x00007fffa1197884 in tevent_common_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:950
  #58 0x00007fffa119ffc0 in std_event_loop_wait (ev=0x154328120, location=0x11db60b50 "../../source3/smbd/server.c:1499") at ../../lib/tevent/tevent_standard.c:141
  #59 0x00007fffa1197978 in _tevent_loop_wait (ev=<optimized out>, location=<optimized out>) at ../../lib/tevent/tevent.c:971
  #60 0x000000011db58c54 in smbd_parent_loop (parent=<optimized out>, ev_ctx=0x154328120) at ../../source3/smbd/server.c:1499
  #61 main (argc=<optimized out>, argv=<optimized out>) at ../../source3/smbd/server.c:2258

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(backported from commit 2eef298ff4c5baf15c7d29c65fb021dbed5b0a93)
[slow@samba.org: changed argument of share_mode_watch_send()]
[slow@samba.org: small context change in vfs_fruit]

s3/locking: add brl_set_modified()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 2772f147c9b13cd2160181c4f7905b54ab765054)

s3/brlock: don't increment current_lock_count if do_lock_fn() failed

Also only assign psmblctx and pblocker_pid if the lock request failed.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 3a0c6e99de4377f44bc29766b6ceb79040caed9f)

s3/brlock: add share_mode_do_locked_brl()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit e17fb732c89f8b34de00904383044de3c4f85bd0)

s3/brlock: add brl_req_set()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit c9c04c7d75dee0c3e6e843b581624a3852042057)

s3/brlock: split out brl_get_locks_readonly_parse()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Pair-Programmed-With: Ralph Boehme <slow@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 94e7cbcc32b73e4d56e7209e04d22d4270a6eb5b)

smbtorture: add test "open-brlock-deadlock"

smbtorture reproducer for bug 15767. As it needs a very specific setup that
can't easily be done in selftest, the test is only executed when manually called
with

  --option=torture:open_brlock_deadlock_timemout=SEC

To prepare the setup for the test set:

  tdb_hash_size:locking.tdb = 1
  tdb_hash_size:brlock.tdb = 1

and remove both tdb from disk which is needed so the TDBs get recreated with the
new hash_size.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 7c60498cee7dca5770d4d1f623c472d585ae9cae)

dbwrap: check for option "tdb_hash_size:DBNAME.tdb" in db_open()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15767

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 7eb135c42d530a16e80e165d9e8e99d920797f12)

VERSION: Bump version up to Samba 4.21.6...

and re-enable GIT_SNAPSHOT.

Signed-off-by: Jule Anger <janger@samba.org>

VERSION: Disable GIT_SNAPSHOT for the 4.21.5 release.

Signed-off-by: Jule Anger <janger@samba.org>

WHATSNEW: Add release notes for Samba 4.21.5.

Signed-off-by: Jule Anger <janger@samba.org>

vfs: Fix Bug 15791, vfs_acl_tdb unlinkat()

Bug: https://bugzilla.samba.org/show_bug.cgi?id=15791
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 93a6d36239dd2ce2b3863945f8b9b59cb6aa911a)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Mon Mar 31 11:14:47 UTC 2025 on atb-devel-224

vfs: Fix a lock order violation in unlinkat_acl_tdb()

unlinkat is called when the share mode record is locked.

Bug: https://bugzilla.samba.org/show_bug.cgi?id=15791
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 93bc238aa91ec8041648d17e11bf235132974eda)

vfs_ceph_new: Add path based fallback for SMB_VFS_FNTIMES

Fallback mechanism was missing in vfs_ceph_fntimes() for path
based call.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15834

Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
Autobuild-User(master): Günther Deschner <gd@samba.org>
Autobuild-Date(master): Mon Mar 17 20:48:55 UTC 2025 on atb-devel-224

(cherry picked from commit dbc48a4cda7489363688bb38f6fa678011fedfaf)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Mon Mar 24 14:01:10 UTC 2025 on atb-devel-224

vfs_ceph_new: Add path based fallback for SMB_VFS_FCHMOD

Fallback mechanism was missing in vfs_ceph_fchmod() for path based call.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15834

Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
(cherry picked from commit 9c019ecf4eae6e6bef48323a0b093e17b0708ee8)

vfs_ceph_new: Add path based fallback for SMB_VFS_FCHOWN

Fallback mechanism was missing in vfs_ceph_fchown() for path based call.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15834

Signed-off-by: Anoop C S <anoopcs@samba.org>
Reviewed-by: Guenther Deschner <gd@samba.org>
(cherry picked from commit abb97683902f50b2a57989f30c0fb53fd3492af9)

s3/lib: fix matching interfaces with multiple assigned IPs

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15823

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Björn Jacke <bjacke@samba.org>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Martin Schwenke <martins@samba.org>
Autobuild-Date(master): Wed Mar 12 01:32:30 UTC 2025 on atb-devel-224

(cherry picked from commit b85f056e7312ea9839b6fda617132fcc956da3c1)

Autobuild-User(v4-21-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-21-test): Thu Mar 20 14:50:08 UTC 2025 on atb-devel-224