s4:dsdb/schema: remember if a backlink attribute is not allowed on class 'top'
Backlink attributes which are not "allowed" in objectClass 'top'
are always possible, but only visible by default based on the
real objectClass.
In order to avoid pay the cost for finding out if a backlink
should be visible or not, we remember a 'bool bl_maybe_invisible'
both on the forward link as well as the backlink dsdb_attribute.
s4:dsdb/tests: let a test to demonstrate the behavior of invisible backlinks
msDS-KeyCredentialLink/msDS-KeyCredentialLink-BL are defined as linked attribute pair,
but msDS-KeyCredentialLink-BL is not defined as allowed on any object class definition,
still it's possible to create msDS-KeyCredentialLink attributes.
msDS-KeyPrincipal/msDS-KeyPrincipalBL are also defined as linked attribute pair
and msDS-KeyPrincipalBL is only allowed on object class 'user', but it's possible
to create msDS-KeyPrincipal values pointing to non 'user' objects.
The result is that 'user' objects have a visible msDS-KeyPrincipalBL, but
the others don't have msDS-KeyPrincipalBL visible, by default.
The backlinks are always visible if the backlink attributes are
explicitly requested.
s4:dsdb/tests: let linked_attributes.py use a container as testbase
The tests don't require an organizationalUnit as parent object
and it makes it possible to create msDS-KeyCredential objects
within the container in the next step.
python:join: run domain adprep as part of join_provision_own_domain()
This is currently unused as we don't support more than one
domain per forest, but it will help it future.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Wed Mar 22 23:05:39 UTC 2023 on atb-devel-224
python/samba: let get_domain_descriptor() include adprep 2016 ACEs
We need to make sure a new provision as well as dbcheck
--reset-well-known-acls include acls used by adprep 2016,
otherwise we would undo the adprep result.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
domain_update: implement updates 82-89 in order to reach the latest w2016 level
I implemented them by looking at
source4/setup/adprep/WindowsServerDocs/Domain-Wide-Updates.md.unused
and looking at a network capture where a Windows 2022 joins an
Windows 2008R2 domain.
The strange thing is that Windows (tested with server 2022) uses c81fc9cc-0130-f4d1-b272-634d74818133 for update 83, while
Domain-Wide-Updates.md and a fresh installation use c81fc9cc-0130-4fd1-b272-634d74818133. In order to match a fresh
installation we use c81fc9cc-0130-4fd1-b272-634d74818133.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
setup/adprep: import the latest {Domain-Wide,Forest-Wide,Read-Only-Domain-Controller,Schema}-Updates.md
We have Domain-Wide-Updates.md and Read-Only-Domain-Controller-Updates.md only
for completeness, they are not parsed/used yet, so we added .unused in
order to avoid confusion in future.
Initially I tried to go with an ms_domain_updates_markdown.py,
but it is easier to add the current updates by hand to
domain_update.py, which will follow in the next commits.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
domain_update: remove useless searches to '(objectClass=samDomain)'
samDomain is an auxiliary class of domainDNS, so we'll handle them
in the search for domainDNS anyway. In addition searches for auxiliary
classes will never be found in searches.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
blackbox/dbcheck: also run currently unused dbcheck_reset_well_known_acls
This makes sure that we detect if dbcheck --reset-well-known-acls
tries to reset to unexpected values, which we expect to currect in
recent provisions.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Joseph Sutton [Mon, 20 Mar 2023 20:43:01 +0000 (09:43 +1300)]
s4:kdc: Fix typo
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Wed Mar 22 19:36:28 UTC 2023 on atb-devel-224
Joseph Sutton [Mon, 20 Mar 2023 02:16:21 +0000 (15:16 +1300)]
s4:kdc: Split samba_kdc_get_pac_blobs() into smaller functions
Instead of having one large function that returns every PAC blob, we now
have a more manageable assortment of smaller functions that each return
one blob.
That gives us more fine-grained handling of PAC blobs, with callers now
able to procure only the specific blobs that they need.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Joseph Sutton [Mon, 20 Mar 2023 02:02:53 +0000 (15:02 +1300)]
s4:kdc: Don't modify cached user_info_dc SIDs
samba_kdc_get_pac_blobs() passes a pointer to a user_info_dc structure
obtained from samba_kdc_get_user_info_from_db() into
samba_add_asserted_identity(). The latter function modifies the SIDs of
the user_info_dc structure in order to add the Asserted Identity SID,
but samba_kdc_get_user_info_from_db() actually caches that structure
internally, meaning that subsequent calls will return the modified
structure.
We should not modify cached SIDs, so have
samba_kdc_get_user_info_from_db() return a pointer to constant data, and
copy the returned array of SIDs before adding the Asserted Identity SID.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Joseph Sutton [Thu, 16 Mar 2023 22:57:09 +0000 (11:57 +1300)]
s4:kdc: Don't check PAC-OPTIONS claims-supported bit
Windows only consults the PAC-OPTIONS claims bit to find out whether or
not to add claims to the PAC if the ClaimsCompIdFASTSupport option is
set to 1. If this option is set to 2 or 3, the bit is ignored and claims
are always added.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
samba-tool: rewrite dsacl.py to use the new sd_utils helpers
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Wed Mar 22 15:57:15 UTC 2023 on atb-devel-224
They better represent what they are doing, we keep dacl_add_ace()
as wrapper of dacl_prepend_aces() in order to let existing callers
work as before.
In future it would be good to have a dacl_insert_aces() that
would canonicalize the ace order before storing, but that a task
for another day.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
This is a more generic api that can be re-used in other places
as well in future. It operates on a security descriptor object instead of
SDDL.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
This uses ndr_pack/unpack in order to create a deep copy
of the given object.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
py_security: allow idx argument to descriptor.[s|d]acl_add()
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
libcli/security: prepare security_descriptor_acl_add() to place the ace at a position
Often it is important to insert an ace at a specific position in the
ACL. As a default we still append by default by using -1, which is the
generic version of passing the number of existing aces.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
lib/ldb-samba: let ldif_read_ntSecurityDescriptor() only try sddl if isupper()
Trying ndr_pull_security_descriptor on SDDL produces just strange
debug messages, which can cause confusion.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz>
Joseph Sutton [Tue, 15 Mar 2022 01:01:13 +0000 (14:01 +1300)]
libcli/security: Reorder SDDL access flags table to match Windows
This means that encoding an ACE in string form will now match Windows.
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Mar 21 01:19:16 UTC 2023 on atb-devel-224
David Mulder [Thu, 16 Mar 2023 21:31:33 +0000 (15:31 -0600)]
gpupdate: Use winbind separator in PAM Access Policies
Signed-off-by: David Mulder <dmulder@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Mon Mar 20 20:20:41 UTC 2023 on atb-devel-224
In future we need to pass in 3 sids to sddl_encode()
Once we pass in a machine_sid from the caller we need to
have a test on a Windows member if the .machine_rid values
really belong to the local machine sid.
At least [MS-DTYP] 2.4.2.4 Well-Known SID Structures
pretents "LA" and "LG" are relative to the local machine sid.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Mon Mar 20 10:53:41 UTC 2023 on atb-devel-224
We should walk the sid_codes array just once.
This makes further changes easier...
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
In future we'll need more than 'domain_sid' in order
to do the correct transition of SDDL to/from security_descriptor.
In the end we most likely add an
sddl_transition_{create,encode,decode}() api in order
to allow the caller to create an sddl_transition_state
once and then pass it to multiple calls to encode/decode.
Signed-off-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Andrew Bartlett <abartlet@samba.org> Reviewed-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Joseph Sutton [Thu, 16 Mar 2023 20:04:51 +0000 (09:04 +1300)]
s4:kdc: Add client claims blob if it is present
Until we support claims we just return an empty blob,
that matches what Windows is doing without defined claims.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Mon Mar 20 01:25:07 UTC 2023 on atb-devel-224
Joseph Sutton [Thu, 16 Mar 2023 19:02:24 +0000 (08:02 +1300)]
s4:kdc: Refactor PAC handling
It's getting unwieldy adding new PAC buffer types when each one has to
have its own handling. It also makes the possibility of mistakes more
likely.
Add a new container, 'struct pac_blobs', containing the types of PAC
buffers in a given PAC, with an index for quick access to the types we
support specifically. We can add new blobs (overriding existing ones) by
calling pac_blobs_add_blob(), and override certain blobs that must be
present with pac_blobs_replace_existing().
This removes the need to have a complicated 'switch' statement with
different logic for each PAC buffer type, or a dozen index variables.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Joseph Sutton [Wed, 2 Nov 2022 01:57:03 +0000 (14:57 +1300)]
wafsamba: Remove unused configure check
This check would trigger compiler warnings due to the extra argument
passed to eprintf(). HAVE__VA_ARGS__MACRO isn't used anywhere, so we can
remove the check.
Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>