VERSION: Disable GIT_SNAPSHOT for the 4.17.6 release.

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

WHATSNEW: Add release notes for Samba 4.17.6.

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

s3: smbd: Fix fsp/fd leak when looking up a non-existent stream name on a file.

When open_stream_pathref_fsp() returns
NT_STATUS_OBJECT_NAME_NOT_FOUND, smb_fname_rel->fsp
has been set to NULL, so we must free base_fsp separately
to prevent fd-leaks when opening a stream that doesn't
exist.

Remove knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15314

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Fri Mar  3 16:37:27 UTC 2023 on atb-devel-224

(cherry picked from commit 3f84a6df4546e0f1e62dfbcd0b823ea29499a787)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Wed Mar  8 10:11:41 UTC 2023 on sn-devel-184

s3: tests: Add new test_stream_dir_rename.sh test.

Shows we are leaking an fsp/fd if we request a non-existent stream on a file.
This then causes rename of a directory containing the file to be denied, as
it thinks we have an existing open file below it.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15314

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit c54bec26ad23b0121b2ddfbf04bc81050f27e6e1)

s3: provision: Add new streams_xattr_nostrict share - needs "strict rename = no".

The bug we're testing for needs "strict rename = no" (the default),
but the existing streams_xattr share uses "strict rename = yes" from
the [global] section.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15314

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 5a3db5105bd8360b245cd35810002740ccff605c)

rpcd: With npa->need_idle_server we can have more than 256 servers

Before this patch the worker-status cut the worker index such that
samba-dcerpcd could not properly update status of the surplus rpc
daemons. This could lead to those daemons to stay around forever,
samba-dcerpcd will never notice they are idle and can exit.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15310

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): Mon Mar  6 22:35:00 UTC 2023 on atb-devel-224

(cherry picked from commit a1780ed8d1b46e4760319b27a4978e7ce7a1df80)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Tue Mar  7 11:16:46 UTC 2023 on sn-devel-184

rpcd: Do blocking connects to local pipes

We don't have real async callers yet, and this is the simplest way to
fix our missing light-weight deterministic async fallback mechanism.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15310

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 0ab7b84ccbd383bb2f696ce258438d4af57fe6f0)

rpcd: Increase listening queue

Allow more waiters under load.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15310

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit f23eb1b3b728d7a13add8c6614d9992aad2c3653)

torture3: test rpc scalability

With

smbtorture3 //127.0.0.1/ipc\$ rpc-scale -N 50 -o 1000

I am able to immediately trigger bug 15130.

Not running by default, this is a pure load test.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15310

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(back-ported from commit 86e95b57d6848482dc73c624c2e8d2bdb79c1d21)

librpc: Remove unused sync rpc_transport_np_init()

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit f3ce9970002459a069344ca5519fb91feacb4a6e)

librpc: Make rpc_pipe_open_np() public and async

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 07ebf97a74fb5c0d0504e76c50f3aca8257dab1f)

lib:util: File descriptor being closed repeatedly.

In file_load()/file_lines_load(), the file's fd is obtained using
open(), and in fd_load() the fd is converted to a FILE* using
fdopen(). However, after fclose(), the fd is closed again using
close().

Bug: https://bugzilla.samba.org/show_bug.cgi?id=15311
Signed-off-by: baixiangcpp baixiangcpp@gmail.com
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Thu Feb 16 12:13:05 UTC 2023 on atb-devel-224

(cherry picked from commit 206dcf7d426e9e85c896c54839008e194d9a2824)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Thu Feb 23 13:37:27 UTC 2023 on sn-devel-184

vfs_ceph: use fsp_get_pathref_fd in ceph fstatat and close vfs calls

Replace fsp_get_io_fd with fsp_get_pathref_fd as these calls do use
pathref fsps. fsp_get_io_fd asserts that the fsp is not pathref and
asserts (on a debug build) or returns -1 (non debug build).

Prior to these changes running ls on the root of the share failed.
Logging from the failure case:
```
openat_pathref_fsp: smb_fname [.]
openat_pathref_fullname: smb_fname [.]
fsp_new: allocated files structure (1 used)
file_name_hash: //. hash 0x3dfcc1c2
check_reduced_name: check_reduced_name [.] [/]
cephwrap_realpath: [CEPH] realpath(0x55604da9a030, .) = //.
check_reduced_name realpath [.] -> [//.]
check_reduced_name: . reduced to //.
cephwrap_openat: [CEPH] openat(0x55604da9a030, ., 0x55604da81f00, 133120, 0)
cephwrap_openat: [CEPH] open(...) = 10
cephwrap_fstat: fsp_get_io_fd: fsp [.] is a path referencing fsp
[CEPH] fstat(0x55604da9a030, -1)
fsp_get_io_fd: fsp [.] is a path referencing fsp
cephwrap_fstat: [CEPH] fstat(...) = -9
fd_openat: name ., flags = 04000 mode = 00, fd = 10.  NT_STATUS_INVALID_HANDLE
openat_pathref_fullname: Opening pathref for [.] failed: NT_STATUS_INVALID_HANDLE
```

This change also seems to match the recommendations in the `When to use
fsp_get_io_fd() or fsp_get_pathref_fd()` section of The_New_VFS.txt
document.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15307

Signed-off-by: John Mulligan <jmulligan@redhat.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Gunther Deschner <gdeschne@redhat.com>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Mon Feb 13 20:04:38 UTC 2023 on atb-devel-224

(cherry picked from commit 54a8da864071e28eb6297b872dcb57fb9b171f33)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Wed Feb 15 14:21:41 UTC 2023 on sn-devel-184

mdssvc: fix kMDScopeArray parsing

In macOS Ventura marshalling of kMDScopeArray in the "openQueryWithParams"
request has changed from

  string: kMDScopeArray
  sl_array_t(#1): {
    string: /foo/bar
  }

to:

  string: kMDScopeArray
  sl_array_t(#1): {
    sl_array_t(#1): {
      string: /foo/bar
    }
  }

With this patch we check both encodings. Bug fixed according to user feedback.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15299
RN: Spotlight doesn't work with latest macOS Ventura

Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit 1dd0cd2f4e644ad43b0e7ee2aaae19799e859585)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Fri Feb 10 10:43:35 UTC 2023 on sn-devel-184

s4-drsuapi: Give an error that matches windows on destination_dsa_guid lookup failure

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Jan 31 13:43:54 UTC 2023 on atb-devel-224

(cherry picked from commit 0f2978bbc0ed5b65d75c20472650a749643312e7)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Fri Feb  3 10:28:42 UTC 2023 on sn-devel-184

s4-drsuapi: Clarify role of drs_security_access_check_nc_root()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 1838f349c94b878de1740af35351a2e8e0c8cffb)

s4-rpc_server: Pre-check destination_dsa_guid in GetNCChanges for validity

This allows our new tests to pass as these need to be checked first.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 115a3a10440f44ba11029be5ae3a05534a7b98c0)

s4-drsuapi: Use samdb_get_ntds_obj_by_guid() to find RODC in REPL_SECRET

We need to find the RODC per the destination_dsa_guid to mark the secrets as
having been replicated, and by using samdb_get_ntds_obj_by_guid() we are stricter
in the checks, as the RODC has to be the right objectClass (nTDSDSA) and under
the CN=Configuration partition.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 09ec6a1db2d3b831548bf7d66475c486be29b1d1)

s4-dsdb: Require that the NTDS object is an nTDSDSA objectclass

This should avoid a user being able to specify the GUID of a different
type of object.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit adb776149e5ac0eb346992775610627106e1a986)

s4-dsdb: Split samdb_get_ntds_obj_by_guid() out of samdb_is_rodc()

This will allow the logic here to be tighened up and shared
in the next few commits.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit d5a2af3feae98057ba29de444d308d499d633941)

s4-rpc_server/drsuapi: Return correct error code for an invalid DN to EXOP_REPL_OBJ/EXOP_REPL_OBJ

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit cbe18353d8d7b2a35b965e4fc8c895ac497e67e8)

s4-drs: Make drs_ObjectIdentifier_to_dn() safer and able to cope with DummyDN values

We want to totally ignore the string DN if there is a GUID,
as clients like "Microsoft Azure AD connect cloud sync" will
set a literal "DummyDN" string.

RN: Use of the Azure AD Connect cloud sync tool is now supported for
password hash synchronisation, allowing Samba AD Domains to synchronise
passwords with this popular cloud environment.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 73f3ece8b2b44ac4b3323a08fb969f29bf2b0380)

s4-dsdb: rework drs_ObjectIdentifier_to_dn() into drs_ObjectIdentifier_to_dn_and_nc_root()

This make this funciton the gatekeeper between the wire format and the
internal struct ldb_dn, checking if the DN exists and which NC
it belongs to along the way, and presenting only a DB-returned
DN for internal processing.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit aee2039e63ceeb5e69a0461fb77e0f18278e4dc4)

s4-rpc_server/drsuapi: Use dsdb_normalise_dn_and_find_nc_root()

This reuses the search done for dsdb_find_nc_root() to normalise the DN.

This will allow a GUID-input DN to be compared safely with a RID Manager DN
or Naming Context.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit e96dfc74b3ece40fe64a33aa8b8d810b576982bd)

s4-dsdb: Add dsdb_normalise_dn_and_find_nc_root() around dsdb_find_nc_root()

Reuse the search done for dsdb_find_nc_root() to normalise the DN.

This will allow a GUID-input DN to be compared safely with a RID Manager DN
or Naming Context.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 8e1122420efd11a91aa1c5d60c0cc8fd9ffaf157)

s4-dsdb: Add better debugging to dsdb_objects_have_same_nc()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 0f501b2316af6568003e520848c1ec80c286fd36)

s4-dsdb: Make dsdb_find_nc_root() first try and use DSDB_CONTROL_CURRENT_PARTITION_OID

This allows lookup of a DN with a GUID only or GUID and string,
possibly not yet in the database, yet still getting the correct result.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit d0444be4b74bdad6a731bc5fcf86da6142b03539)

s4-dsdb: Schedule SD propegation only after successful rename

This avoids needing to anticipate errors that the rename might give
while allowing the dsdb_find_nc_root() routine to become stricter.

The problem is that dsdb_find_nc_root() will soon do a real search and
so fail more often, but these failures will give "wrong" error codes.

We do not need to do this work if the operation fails, so put this in
the callback.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 7032b86cd5c1456318558ed95f8890e353117ced)

s4-selftest/drs: Confirm GetNCChanges REPL_SECRET works with a DummyDN and real GUID

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 7c43388576f768db564aaf15a47d3f9ce5796fb3)

s4-selftest/drs: Confirm GetNCChanges full replication works with a DummyDN and real GUID

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 539221dda33f03a1abf5ee5f3153db0fe1a9bfe6)

s4-selftest/drs: Confirm GetNCChanges REPL_OBJ works with a DummyDN and real GUID

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 70faccae6d595056174af8d63b3437c9fe3805aa)

s4-selftest/drs Allow re-run of DRS tests after failed cleanup

Using a random base is a useful start, even if the better solution also includes a self.AddCleanup()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 3204d1350b21704474e577cb5f3f2439b673c421)

s4-selftest/drs Allow some DRS tests to operate against an IP

This is not comprehensive, but makes some manual test runs easier by
avoiding the need for DNS names to resolve.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit a150a2dcb1fc7fc7f606838de17ad4d3e6072bda)

s4-selftest/drs Add test of expected return code for invaid DNs in GetNCChanges

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit bee45e6b29b97e0cab19a9c3cf692d9a7585a717)

s4-dsdb: Add tests of SamDB.get_nc_root()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=10635

Signed-off-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 2c7bb58703c1fa26782ac6959ea7d81fccf3905c)

s3/lib: Prevent use after free of messaging_ctdb_fde_ev structs

In a cluster setup samba-bgqd async callback
cups_pcap_load_async can access messaging_ctdb_fde_ev associated
with already destructed global_ctdb_ctx_destructor

==26053== Invalid read of size 8
==26053==    at 0x71692E1: messaging_ctdb_fde_ev_destructor (messages_ctdb.c:181)
==26053==    by 0x40B2309: _tc_free_internal (talloc.c:1158)
==26053==    by 0x40B3539: _tc_free_children_internal (talloc.c:1669)
==26053==    by 0x40B24C4: _tc_free_internal (talloc.c:1184)
==26053==    by 0x40B3539: _tc_free_children_internal (talloc.c:1669)
==26053==    by 0x40B24C4: _tc_free_internal (talloc.c:1184)
==26053==    by 0x40B2685: _talloc_free_internal (talloc.c:1248)
==26053==    by 0x40B3963: _talloc_free (talloc.c:1792)
==26053==    by 0x4056BCA: tevent_req_received (tevent_req.c:301)
==26053==    by 0x405673D: tevent_req_destructor (tevent_req.c:135)
==26053==    by 0x40B2309: _tc_free_internal (talloc.c:1158)
==26053==    by 0x40B3539: _tc_free_children_internal (talloc.c:1669)
==26053==    by 0x40B24C4: _tc_free_internal (talloc.c:1184)
==26053==    by 0x40B2685: _talloc_free_internal (talloc.c:1248)
==26053==    by 0x40B3963: _talloc_free (talloc.c:1792)
==26053==    by 0x1384EF: cups_pcap_load_async (print_cups.c:507)
==26053==    by 0x13894B: cups_cache_reload (print_cups.c:602)
==26053==    by 0x1373AE: pcap_cache_reload (pcap.c:140)
==26053==    by 0x1369D2: register_printing_bq_handlers (queue_process.c:323)
==26053==    by 0x122AD6: main (samba-bgqd.c:316)
==26053==  Address 0xed64d48 is 120 bytes inside a block of size 128 free'd
==26053==    at 0x4C370EB: free (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==26053==    by 0x40B25E1: _tc_free_internal (talloc.c:1222)
==26053==    by 0x40B2685: _talloc_free_internal (talloc.c:1248)
==26053==    by 0x40B3963: _talloc_free (talloc.c:1792)
==26053==    by 0x71691F6: messaging_ctdb_destroy (messages_ctdb.c:141)
==26053==    by 0x7169C21: msg_ctdb_ref_destructor (messages_ctdb_ref.c:142)
==26053==    by 0x40B2309: _tc_free_internal (talloc.c:1158)
==26053==    by 0x40B3539: _tc_free_children_internal (talloc.c:1669)
==26053==    by 0x40B24C4: _tc_free_internal (talloc.c:1184)
==26053==    by 0x40B2685: _talloc_free_internal (talloc.c:1248)
==26053==    by 0x40B3963: _talloc_free (talloc.c:1792)
==26053==    by 0x4157380: messaging_reinit (messages.c:646)
==26053==    by 0x416C01E: reinit_after_fork (util.c:488)
==26053==    by 0x13844C: cups_pcap_load_async (print_cups.c:498)
==26053==    by 0x13894B: cups_cache_reload (print_cups.c:602)
==26053==    by 0x1373AE: pcap_cache_reload (pcap.c:140)
==26053==    by 0x1369D2: register_printing_bq_handlers (queue_process.c:323)
==26053==    by 0x122AD6: main (samba-bgqd.c:316)
==26053==  Block was alloc'd at
==26053==    at 0x4C346A4: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==26053==    by 0x40B1989: __talloc_with_prefix (talloc.c:783)
==26053==    by 0x40B1B23: __talloc (talloc.c:825)
==26053==    by 0x40B1ECC: _talloc_named_const (talloc.c:982)
==26053==    by 0x40B49C3: _talloc_zero (talloc.c:2421)
==26053==    by 0x7168E68: messaging_ctdb_init (messages_ctdb.c:93)
==26053==    by 0x716979D: messaging_ctdb_ref (messages_ctdb_ref.c:75)
==26053==    by 0x415702A: messaging_init_internal (messages.c:563)
==26053==    by 0x41572FD: messaging_init (messages.c:622)
==26053==    by 0x4163ED3: global_messaging_context (global_contexts.c:62)
==26053==    by 0x12273B: main (samba-bgqd.c:271)
==26053==

Bug: https://bugzilla.samba.org/show_bug.cgi?id=15293

Signed-off-by: Noel Power <npower@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 7a880ef52dfc85ed2f674250b5baf5109f8d4691)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Fri Jan 27 09:45:10 UTC 2023 on sn-devel-184

VERSION: Bump version up to Samba 4.17.6...

and re-enable GIT_SNAPSHOT.

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

VERSION: Disable GIT_SNAPSHOT for the 4.17.5 release.

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

WHATSNEW: Add release notes for Samba 4.17.5.

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

lib/replace - add extra check to bsd_attr_list

The FreeBSD extattr API may return success and truncated
namelist. We need to check for this in bsd_attr_list to
ensure that we don't accidentally read off the end of the
buffer. In the case of a truncated value, the pascal
strings for attr names will reflect the lengths as if
the value were not truncated. For example:
`58DosStrea`

In case of short read we now set error to ERANGE and
fail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15271

Signed-off-by: Andrew Walker <awalker@ixsystems.com>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Mon Jan  2 14:27:23 UTC 2023 on sn-devel-184

(cherry picked from commit 01cdc5e00be78a51f0766634cc7fe50de2088203)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Mon Jan 23 10:27:59 UTC 2023 on sn-devel-184

s3: smbd: Always use metadata_fsp() when processing fsctls.

Currently all fsctls we implement need the base fsp, not
an alternate data stream fsp. We may revisit this later
if we implement fsctls that operate on an ADS.

Remove knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15236

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Andrew Walker <awalker@ixsystems.com>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Mon Nov 14 18:13:31 UTC 2022 on sn-devel-184

(cherry picked from commit fa4eba131b882c3858b28f5fd9864998e19a4510)

s3: smbd: Add test to show smbd crashes when doing an FSCTL on a named stream handle.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15236

Signed-off-by: Andrew Walker <awalker@ixsystems.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit abc4495e4591964bb4625c2669a1f84213faab77)

s3:auth: call wbcFreeMemory(info) in auth3_generate_session_info_pac()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15286

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
(cherry picked from commit e27084f5d8c3a151c5d0b266118f0d71b641dc85)

CVE-2022-38023 s3:rpc_server/netlogon: Avoid unnecessary loadparm_context allocations

After s3 and s4 rpc servers merge the loadparm_context is available in
the dcesrv_context structure.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Mon Jan  9 15:17:14 UTC 2023 on sn-devel-184

(cherry picked from commit 56837f3d3169a02d0d92bd085d9c8250415ce29b)

CVE-2022-38023 docs-xml/smbdotconf: The "server schannel require seal[:COMPUTERACCOUNT]" options are also honoured by s3 netlogon server.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 02fba22b8c9e9b33ab430555ef45500c45eaa9d1)

CVE-2022-38023 s3:rpc_server/netlogon: Check for global "server schannel require seal"

By default we'll now require schannel connections with privacy/sealing/encryption.

But we allow exceptions for specific computer/trust accounts.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit a0b97e262318dc56fe663da89b0ee3172b2e7848)

CVE-2022-38023 s3:rpc_server/netlogon: make sure all _netr_LogonSamLogon*() calls go through dcesrv_netr_check_schannel()

Some checks are also required for _netr_LogonSamLogonEx().

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit ca07f4340ce58a7e940a1123888b7409176412f7)

CVE-2022-38023 s3:rpc_server/netlogon: Use dcesrv_netr_creds_server_step_check()

After s3 and s4 rpc servers merge we can avoid duplicated code.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 25300d354c80995997d552581cd91dddaf4bbf48)

CVE-2022-38023 s4:rpc_server/netlogon: Move schannel and credentials check functions to librpc

Will be used later by s3 netlogon server.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 121e7b0e39478c5291100652ac92c263f406076b)

CVE-2022-38023 s4:rpc_server:wscript: Reformat following pycodestyle

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit d9e6b490db3ead7e79bb3ff0c1f9ef8ab8bdc65b)

CVE-2022-38023 selftest:Samba3: avoid global 'server schannel = auto'

Instead of using the generic deprecated option use the specific
server require schannel:COMPUTERACCOUNT = no in order to allow
legacy tests for pass.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 3cd18690f83d2f85e847fc703ac127b4b04189fc)

CVE-2022-38023 s3:rpc_server/netlogon: 'server schannel != yes' warning to dcesrv_interface_netlogon_bind

Follow s4 netlogon server changes and move the checks to the RPC bind
hook. Next commits will remove the s3 netr_creds_server_step_check()
function.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15240

Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
(cherry picked from commit 8141eae47aad849741beb138fae866c772e4ec4c)

s3: smbd: Tweak openat_pathref_dirfsp_nosymlink() to NULL out fsp->fsp_name after calling fd_close() on intermediate directories, rather than before.

vfs_virusfilter expects a non-NULL fsp->fsp_name to use for printing debugs
(it always indirects fsp->fsp_name). vfs_fruit also does the same, so would
also crash in fruit_close() with 'debug level = 10' and vfs_default:VFS_OPEN_HOW_RESOLVE_NO_SYMLINKS = no
set (we don't test with that which is why we haven't noticed
this before).

Remove knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15283

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Fri Jan 13 08:33:47 UTC 2023 on sn-devel-184

(cherry picked from commit 3d3d01cda8d3a6d0d18d1b808aa9414e71d56062)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Mon Jan 16 10:49:14 UTC 2023 on sn-devel-184

selftest: Show vfs_virusscanner crashes when traversing a 2-level directory tree.

Modify check_infected_read() test to use a 2-level deep
directory.

We must have vfs_default:VFS_OPEN_HOW_RESOLVE_NO_SYMLINKS = no
set on the virusscanner share as otherwise the openat flag
shortcut defeats the test.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15283

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
(cherry picked from commit c844bff3eca336547c6cedfeeb03adda4eed57c6)

s4: libcli: Ignore errors when getting A records after fetching AAAA records.

The target may only be available over IPv6.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15226

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Wed Nov  9 20:34:07 UTC 2022 on sn-devel-184

(cherry picked from commit 10537a89bb0b461ba31d614b7c9ed56a842422e7)

s3: smbd: In synthetic_pathref() change DBG_ERR -> DBG_NOTICE to avoid spamming the logs.

Can easily be seen by doing make test TESTS=fruit
and looking in st/nt4_dc/smbd_test.log.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15210

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Wed Nov 16 06:00:56 UTC 2022 on sn-devel-184

(cherry picked from commit f0ca9546102acf09f1834c03f8907ed26bfc80f8)

s3: smbd: Cause SMB2_OP_FLUSH to go synchronous in a compound anywhere but the last operation in the list.

Async read and write go synchronous in the same case,
so do the same here.

Remove knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15172

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Autobuild-User(master): Ralph Böhme <slow@samba.org>
Autobuild-Date(master): Thu Nov 17 05:55:42 UTC 2022 on sn-devel-184

(cherry picked from commit 26adf3344337f4e8d5d2107e6ba42e5ea7656372)

s3: smbd: Add utility function smbd_smb2_is_last_in_compound().

Not yet used. Returns true if we're processing the last SMB2 request in a
compound.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15172

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit e668c3a82cd566b405c976d45659dd79786948de)

s4: torture: Add an async SMB2_OP_FLUSH + SMB2_OP_FLUSH test to smb2.compound_async.

Shows we fail sending an SMB2_OP_FLUSH + SMB2_OP_FLUSH
compound if we immediately close the file afterward.

Internally the flushes go async and we free the req, then
we process the close. When the flushes complete they try to access
already freed data.

Extra test which will allow me to test when the final
component (flush) of the compound goes async and returns
NT_STATUS_PENDING.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15172

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 6f149dfd9d8d2619a9e18975ebcf5e69df2b7766)

s4: torture: Add an async SMB2_OP_FLUSH + SMB2_OP_CLOSE test to smb2.compound_async.

Shows we fail sending an SMB2_OP_FLUSH + SMB2_OP_CLOSE
compound. Internally the flush goes async and
we free the req, then we process the close.
When the flush completes it tries to access
already freed data.

Found using the Apple MacOSX client at SNIA SDC 2022.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15172

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 17a110c1b58196eb8ecf3c76eb97e8508976c544)

nsswitch:libwbclient - fix leak in wbcCtxPingDc2

Memory allocated for response is never freed.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15164

Signed-off-by: Andrew Walker <awalker@ixsystems.com>
Reviewed-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Christof Schmitt <cs@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Sep  6 20:10:17 UTC 2022 on sn-devel-184

(cherry picked from commit aa9f3a2da97ae13cce3e50fe3d58f143200e9a17)

s3: libsmbclient: Fix smbc_getxattr() to return 0 on success.

Remove knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14808

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: David Mulder <dmulder@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Nov  1 18:31:22 UTC 2022 on sn-devel-184

(cherry picked from commit bdbb38d16c8eaff33484bb747efa639c4d8e7f35)

s4: torture: Show return value for smbc_getxattr() is incorrect (returns >0 for success, should return zero).

Add torture test to show smbc_getxattr() should return -1 on
failure, 0 on success.

Add knownfail.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=14808

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: David Mulder <dmulder@samba.org>
(cherry picked from commit 74636dfe24c15677261fc40c0a4ec62404898cf4)

s3:smbstatus: go to cmdline_messaging_context_free

If the locking.tdb is not found,
(for example, fresh new installed samba server is not running yet)
smbstatus utility would exit earlier,
and lock files are left behind in the directory 'msg.sock' and 'msg.lock'.
Consider that a script to run smbstatus utility in a loop,
this might result in used space slowly growing-up on the underlying filesystem.
Since the samba server is not running yet,
there is no cleanupd daemon could delete these files to reclaim space.

Supposed to use 'ret = 0; goto done;' instead of exit(0),
this would go through the cmdline_messaging_context_free() which deletes
the lock files in the directory msg.sock and msg.lock before smbstatus
utility is exiting.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15282

Signed-off-by: Jones Syue <jonessyue@qnap.com>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Wed Jan 11 17:08:10 UTC 2023 on sn-devel-184

(cherry picked from commit de5d31f452b2445bd92b1746efb05aa096716af8)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Fri Jan 13 11:56:59 UTC 2023 on sn-devel-184

source3/wscript: Remove implicit int and implicit function declarations

This should fix the remaining C89isms in these configure checks.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15281

Signed-off-by: Florian Weimer <fweimer@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 7779050a6765a37cc0c6438309e512782220dcb3)

source3/wscript: Fix detection of major/minor macros

These macros are only available via <sys/sysmacros.h> as of glibc
commit e16deca62e16f645213dffd4ecd1153c37765f17 ("[BZ #19239] Don't
include sys/sysmacros.h from sys/types.h."), which went into
glibc 2.28.

This is different from the usual C99 cleanups because it changes
the configure check result with existing compilers that usually
accept implicit function declarations.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15281

Signed-off-by: Florian Weimer <fweimer@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit d0ee9d5a43072ecbd37327f5dc936c546f97ba34)

buildtools/wafsamba: Avoid calling lib_func without a prototype

This is a backport of commit f4c0a750d4adebcf2342a44e85f04526c34
("WAF: Fix detection of linker features")
to buildtools/wafsamba/samba_conftests.py.  It fixes the check for
rpath support with compilers in strict C99 mode.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15281

Signed-off-by: Florian Weimer <fweimer@redhat.com>
Reviewed-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 75db84b1e59a42737343445c43c41494460c89f0)

s4:lib/messaging: fix interaction between imessaging_context_destructor and irpc_destructor

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15280

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 0d096931196524a2d1bf59470bc629dc9231131e)

s3:rpc_server/srvsvc: make sure we (re-)load all shares as root.

This fixes a regression in commit f03665bb7e8ea97699062630f2aa1bac4c5dfc7f

The use of reload_services() has a lot of side effects, e.g. reopen of
log files and other things, which are only useful in smbd, but not in rpcd_classic.
It was also unloading the user and registry shares we loaded a few lines
above.

We need to do all (re-)loading as root, otherwise we won't be able
to read root only smb.conf files, access registry shares, ...

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15243
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15266

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Walker <awalker@ixsystems.com>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Thu Dec 29 21:14:02 UTC 2022 on sn-devel-184

(cherry picked from commit f28553105be7465026bcc0fcbbed6a1a8c2133dd)

selftest: add samba3.blackbox.registry_share

This demonstrates the regression introduced by
f03665bb7e8ea97699062630f2aa1bac4c5dfc7f, where
registry shares are no longer listed.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15243
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15266

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Walker <awalker@ixsystems.com>
(cherry picked from commit a00c7395fbc7974a61a70ae54ea6ae6349933de2)

testprogs: Add testit_grep_count() helper

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
(cherry picked from commit 55feb593012fc5b24e795a00081666fca740429c)

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15243
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15266

s3: smbd: Strip any leading '\\' characters if the SMB2 DFS flag is set.

MacOS clients send SMB2 DFS pathnames as \server\share\file\name.

Ensure smbd can cope with this by stipping any leading '\\'
characters from an SMB2 packet with the DFS flag set.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15277

NB. The test for this is not back-ported to
4.17 as there are too many changes in the
test infrastructure and supporting client
libraries between 4.17 and master.

Back-ported from c9a6e242d15ee707a2e30f973fd37e80b3225aca.

Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Thu Jan  5 10:01:54 UTC 2023 on sn-devel-184

s3:client: Fix a use-after-free issue in smbclient

Detected by

    make test TESTS="samba3.blackbox.chdir-cache"

with an optimized build or with AddressSanitizer.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15268

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 9c707b4be27e2a6f79886d3ec8b5066c922b99bd)

Autobuild-User(v4-17-test): Jule Anger <janger@samba.org>
Autobuild-Date(v4-17-test): Wed Jan  4 21:23:48 UTC 2023 on sn-devel-184

s3:script: Improve test_chdir_cache.sh

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15268

Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
(cherry picked from commit 0d1961267cd9e8f1158a407c5d135514c363f37e)

s3:params:lp_do_section - protect against NULL deref

iServiceIndex may indicate an empty slot in the ServicePtrs
array. In this case, lpcfg_serivce_ok(ServicePtrs[iServiceIndex])
may trigger a NULL deref and crash. Skipping the check
here will cause a scan of the array in add_a_service() and the
NULL slot will be used safely.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15267

Signed-off-by: Andrew Walker <awalker@ixsystems.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Tue Dec 20 18:49:54 UTC 2022 on sn-devel-184

(cherry picked from commit 5b19288949e97a5af742ff2719992d56f21e364a)

rpc_server:srvsvc - retrieve share ACL via root context

share_info.tdb has permissions of 0o600 and so we need
to become_root() prior to retrieving the security info.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15265

Signed-off-by: Andrew Walker <awalker@ixsystems.com>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Mon Dec 19 20:41:15 UTC 2022 on sn-devel-184

(cherry picked from commit 80c0b416892bfacc0d919fe032461748d7962f05)

ctdb: Fix a use-after-free in run_proc

If you happen to talloc_free(run_ctx) before all the tevent_req's
hanging off it, you run into the following:

==495196== Invalid read of size 8
==495196==    at 0x10D757: run_proc_state_destructor (run_proc.c:413)
==495196==    by 0x488F736: _tc_free_internal (talloc.c:1158)
==495196==    by 0x488FBDD: _talloc_free_internal (talloc.c:1248)
==495196==    by 0x4890F41: _talloc_free (talloc.c:1792)
==495196==    by 0x48538B1: tevent_req_received (tevent_req.c:293)
==495196==    by 0x4853429: tevent_req_destructor (tevent_req.c:129)
==495196==    by 0x488F736: _tc_free_internal (talloc.c:1158)
==495196==    by 0x4890AF6: _tc_free_children_internal (talloc.c:1669)
==495196==    by 0x488F967: _tc_free_internal (talloc.c:1184)
==495196==    by 0x488FBDD: _talloc_free_internal (talloc.c:1248)
==495196==    by 0x4890F41: _talloc_free (talloc.c:1792)
==495196==    by 0x10DE62: main (run_proc_test.c:86)
==495196==  Address 0x55b77f8 is 152 bytes inside a block of size 160 free'd
==495196==    at 0x48399AB: free (vg_replace_malloc.c:538)
==495196==    by 0x488FB25: _tc_free_internal (talloc.c:1222)
==495196==    by 0x488FBDD: _talloc_free_internal (talloc.c:1248)
==495196==    by 0x4890F41: _talloc_free (talloc.c:1792)
==495196==    by 0x10D315: run_proc_context_destructor (run_proc.c:329)
==495196==    by 0x488F736: _tc_free_internal (talloc.c:1158)
==495196==    by 0x488FBDD: _talloc_free_internal (talloc.c:1248)
==495196==    by 0x4890F41: _talloc_free (talloc.c:1792)
==495196==    by 0x10DE62: main (run_proc_test.c:86)
==495196==  Block was alloc'd at
==495196==    at 0x483877F: malloc (vg_replace_malloc.c:307)
==495196==    by 0x488EAD9: __talloc_with_prefix (talloc.c:783)
==495196==    by 0x488EC73: __talloc (talloc.c:825)
==495196==    by 0x488F0FC: _talloc_named_const (talloc.c:982)
==495196==    by 0x48925B1: _talloc_zero (talloc.c:2421)
==495196==    by 0x10C8F2: proc_new (run_proc.c:61)
==495196==    by 0x10D4C9: run_proc_send (run_proc.c:381)
==495196==    by 0x10DDF6: main (run_proc_test.c:79)

This happens because run_proc_context_destructor() directly does a
talloc_free() on the struct proc_context's and not the enclosing
tevent_req's. run_proc_kill() makes sure that we don't follow
proc->req, but it forgets the "state->proc", which is free()'ed, but
later dereferenced in run_proc_state_destructor().

This is an attempt at a quick fix, I believe we should convert
run_proc_context->plist into an array of tevent_req's, so that we can
properly TALLOC_FREE() according to the "natural" hierarchy and not
just pull an arbitrary thread out of that heap.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15269

Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Martin Schwenke <martin@meltin.net>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Thu Oct  6 15:10:20 UTC 2022 on sn-devel-184

(cherry picked from commit 688be0177b04d04709813a02ae6da1e983ac25dd)

VERSION: Bump version up to Samba 4.17.5...

and re-enable GIT_SNAPSHOT.

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

VERSION: Disable GIT_SNAPSHOT for the 4.17.4 release.

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

WHATSNEW: Add release notes for Samba 4.17.4.

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

s4:libnet: correctly handle gnutls_pbkdf2() errors

We should not ignore the error nor should we map
GNUTLS_E_UNWANTED_ALGORITHM to NT_STATUS_WRONG_PASSWORD,
instead we use NT_STATUS_CRYPTO_SYSTEM_INVALID as in most other places
in the same file.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15206

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Björn Baumbach <bbaumbach@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Dec 14 13:35:20 UTC 2022 on sn-devel-184

(cherry picked from commit eb5df255faea7326a7b85c1e7ce5a66119a27c3a)

Autobuild-User(v4-17-test): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(v4-17-test): Wed Dec 14 14:46:02 UTC 2022 on sn-devel-184

s4:libnet: fix error string for failing samr_ChangePasswordUser4()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15206

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Björn Baumbach <bbaumbach@samba.org>
(cherry picked from commit 53d558365161be1793dad78ebcce877c732f2419)

CVE-2022-37966 python:/tests/krb5: call sys.path.insert(0, "bin/python") before any other imports

This allows the tests to be executed without an explicit
PYTHONPATH="bin/python".

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Dec 13 14:06:14 UTC 2022 on sn-devel-184

(similar to commit 987cba90573f955fe9c781830daec85ad4d5bf92)

Autobuild-User(v4-17-test): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(v4-17-test): Wed Dec 14 12:40:42 UTC 2022 on sn-devel-184

CVE-2022-37966 samba-tool: add 'domain trust modify' command

For now it only allows the admin to modify
the msDS-SupportedEncryptionTypes values.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
(cherry picked from commit d1999c152acdf939b4cd7eb446dd9921d3edae29)

CVE-2022-37966 s4:kdc: apply restrictions of "kdc supported enctypes"

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit cca3c024fc514bee79bb60a686e470605cc98d6f)

CVE-2022-37966 param: Add support for new option "kdc supported enctypes"

This allows admins to disable enctypes completely if required.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 36d0a495159f72633f1f41deec979095417a1727)

CVE-2022-37966 param: let "kdc default domain supportedenctypes = 0" mean the default

In order to allow better upgrades we need the default value for smb.conf to the
same even if the effective default value of the software changes in future.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit fa64f8fa8d92167ed15d1109af65bbb4daab4bad)

CVE-2022-37966 param: don't explicitly initialize "kdc force enable rc4 weak session keys" to false/"no"

This is not squashed in order to allow easier backports...

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 7504a4d6fee7805aac7657b9dab88c48353d6db4)

CVE-2022-37966 s4:kdc: announce PA-SUPPORTED-ETYPES like windows.

We need to take the value from the msDS-SupportedEncryptionTypes
attribute and only take the default if there's no value or
if the value is 0.

For krbtgt and DC accounts we need to force support for
ARCFOUR-HMAC-MD5 and AES encryption types and add the related bits
in addtition. (Note for krbtgt msDS-SupportedEncryptionTypes is
completely ignored the hardcoded value is the default, so there's
no AES256-SK for krbtgt).

For UF_USE_DES_KEY_ONLY on the account we reset
the value to 0, these accounts are in fact disabled completely,
as they always result in KRB5KDC_ERR_ETYPE_NOSUPP.

Then we try to get all encryption keys marked in
supported_enctypes, and the available_enctypes
is a reduced set depending on what keys are
actually stored in the database.

We select the supported session key enctypes by the available
keys and in addition based on AES256-SK as well as the
"kdc force enable rc4 weak session keys" option.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=13135
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit fde745ec3491a4fd7b23e053a67093a2ccaf0905)

CVE-2022-37966 python:tests/krb5: test much more etype combinations

This tests work out the difference between
- msDS-SupportedEncryptionTypes value or it's default
- software defined extra flags for DC accounts
- accounts with only an nt hash being stored
- the resulting value in the KRB5_PADATA_SUPPORTED_ETYPES announcement

BUG: https://bugzilla.samba.org/show_bug.cgi?id=13135
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 1dfa91682efd3b12d7d6af75287efb12ebd9e526)

CVE-2022-37966 python:tests/krb5: add better PADATA_SUPPORTED_ETYPES assert message

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit c7c576208960e336da276e251ad7a526e1b3ed45)

CVE-2022-37966 python:tests/krb5: add 'force_nt4_hash' for account creation of KDCBaseTest

This will allow us to create tests accounts with only an nt4 hash
stored, without any aes keys.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 77bd3258f1db0ddf4639a83a81a1aad3ee52c87d)

CVE-2022-37966 python:tests/krb5: ignore empty supplementalCredentials attributes

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit f434a30ee7c40aac4a223fcabac9ddd160a155a5)

CVE-2022-37966 python:tests/krb5: allow ticket/supported_etypes to be passed KdcTgsBaseTests._{as,tgs}_req()

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit d8fd6a22b67a2b3ae03a2e428cc4987f07af6e29)

CVE-2022-37966 python:tests/krb5: fix some tests running against Windows 2022

I'm using the following options:

SERVER=172.31.9.218 DC_SERVER=w2022-118.w2022-l7.base \
SMB_CONF_PATH=/dev/null STRICT_CHECKING=1 \
DOMAIN=W2022-L7 REALM=W2022-L7.BASE \
ADMIN_USERNAME=Administrator ADMIN_PASSWORD=A1b2C3d4 \
CLIENT_USERNAME=Administrator CLIENT_PASSWORD=A1b2C3d4 CLIENT_AS_SUPPORTED_ENCTYPES=28 CLIENT_KVNO=2 \
FULL_SIG_SUPPORT=1 TKT_SIG_SUPPORT=1 FORCED_RC4=1

in order to run these:

python/samba/tests/krb5/as_req_tests.py -v --failfast AsReqKerberosTests
python/samba/tests/krb5/etype_tests.py -v --failfast EtypeTests

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit e0f89b7bc8025db615dccf096aab4ca87e655368)

CVE-2022-37966 s4:libnet: allow python bindings to force setting an nthash via SAMR level 18

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 4ebbe7e40754eeb1c8f221dd59018c3e681ab2ab)

CVE-2022-37966 s4:libnet: add support LIBNET_SET_PASSWORD_SAMR_HANDLE_18 to set nthash only

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 271cd82cd681d723572fcaeed24052dc98a83612)

CVE-2022-37966 s4:libnet: initialize libnet_SetPassword() arguments explicitly to zero by default.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit 9e69289b099b47e0352ef67ef7e6529d11688e9a)

CVE-2022-37966 drsuapi.idl: add trustedDomain related ATTID values

For now this is only for debugging in order to see
DRSUAPI_ATTID_msDS_SupportedEncryptionTypes in the replication meta
data.

BUG: https://bugzilla.samba.org/show_bug.cgi?id=15219
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit f1c5fa28c460f7e011049606b1b9ef96443e5e1f)

CVE-2022-37966 s4:kdc: use the strongest possible keys

BUG: https://bugzilla.samba.org/show_bug.cgi?id=13135
BUG: https://bugzilla.samba.org/show_bug.cgi?id=15237

Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
(cherry picked from commit d7ea197ed1a9903f601030e6466cc822f9b8f794)