IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
This is slightly less efficient, because we no longer keep a cache on
the gensec structures, but much clearer in terms of memory ownership.
Both gensec_session_info() and gensec_session_key() now take a mem_ctx
and put the result only on that context.
Some duplication of memory in the callers (who were rightly uncertain
about who was the rightful owner of the returned memory) has been
removed to compensate for the internal copy.
Andrew Bartlett
Windows seems to always canonicalize the principal in TGS replies.
metze
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue Jul 26 03:25:06 CEST 2011 on sn-devel-104
TODO: check why this is needed.
metze
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Fri Jul 15 12:26:25 CEST 2011 on sn-devel-104
As mit_samba_update_pac_data() doesn't support adding
S4U_DELEGATION_INFO to the pac (and I have no clue how to add that)
we should disable S4U2Proxy until this is implemented.
metze
Autobuild-User: Stefan Metzmacher <metze@samba.org>
Autobuild-Date: Tue Jun 28 20:35:19 CEST 2011 on sn-devel-104
The two error tables need to be combined, but for now seperate the names.
(As the common parts of the tree now use the _common function,
errmap_unix.c must be included in the s3 autoconf build).
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Jun 20 08:12:03 CEST 2011 on sn-devel-104
The iface_count, iface_n_bcast, and load_interfaces functions
conflicted with functions of the same name in source3, so the source4
functions were renamed. Hopefully we can actually wrap one around the
other in future.
Andrew Bartlett
we shouldn't accept bad multi-byte strings, it just hides problems
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Thu Mar 24 01:47:26 CET 2011 on sn-devel-104
By having kdc_tcp_proxy_send/recv(), which just asks any writeable
dc for a reponse blob, we simplify the interaction between
client-local and local-writeable sockets.
This allows us to make kdc_socket, kdc_process_fn_t, kdc_tcp_call and kdc_tcp_socket
private to kdc.c again.
metze
By having kdc_udp_proxy_send/recv(), which just asks any writeable
dc for a reponse blob, we simplify the interaction between
client-local and local-writeable sockets.
This allows us to make kdc_udp_call and kdc_udp_socket private to
kdc.c again.
metze
Bit shifting is non-trivial in C:-)
This
int32_t a = 0x12340000;
uint32_t b = (a >> 16);
results in 0x00001234, but this
int32_t a = 0xEDCB0000;
uint32_t b = (a >> 16);
results in 0xFFFFEDCB, while we expected 0x0000EDCB.
metze
this prevents conflicts with the s3 auth modules. The auth modules in
samba3 may appear in production smb.conf files, so it is preferable to
rename the s4 modules for minimal disruption.
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
thi ensures we are using the header corresponding to the version of
ldb we're linking against. Otherwise we could use the system ldb for
link and the in-tree one for include
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
This changes auth_serversupplied_info into the IDL-defined struct
auth_user_info_dc. This then in turn contains a struct
auth_user_info, which is the only part of the structure that is
mainted into the struct session_info.
The idea here is to avoid keeping the incomplete results of the
authentication (such as session keys, lists of SID memberships etc) in
a namespace where it may be confused for the finalised results.
Andrew Barltett
This context is used in ldb, and ldb modules apparently abort if nesting is not
allowed.
Autobuild-User: Simo Sorce <idra@samba.org>
Autobuild-Date: Mon Feb 7 20:58:02 CET 2011 on sn-devel-104
We now no longer print tickets with a potentially infinite life, and
we report the same life over LSA as we use in the KDC. We should get
this from group policy, but for now it's parametric smb.conf options.
Andrew Bartlett
Thanks to Hongwei Sun for the clear description of the algorithim
involved. Importantly, it isn't possible to remove encryption types
from the list, only to add them over the defaults (DES and
arcfour-hmac-md5, and additional AES for DCs and RODCs).
This changes the behaviour for entries with
msDS-supportedEncryptionTypes: 0, which Angelos Oikonomopoulos
reported finding set by ADUC when attempting to store cleartext
passwords.
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Tue Nov 16 21:24:43 UTC 2010 on sn-devel-104
The KDC sets different flags for the AS-REQ (this is client-depenent)
and the TGS-REQ to determine if the realm should be forced to the
canonical value. If we do this always, or do this never, we get into
trouble, so it's much better to honour the flags we are given.
Andrew Bartlett
We should never get a cross-realm ticket that was not issued by a full
DC, but if someone claims to have such a thing, reject it rather than
segfaulting on the NULL client pointer.
Andrew Bartlett
Autobuild-User: Andrew Bartlett <abartlet@samba.org>
Autobuild-Date: Mon Nov 15 23:59:34 UTC 2010 on sn-devel-104
The PAC was being regenerated on all normal DCs, because they don't
have a msDS-SecondaryKrbTgtNumber attribute. Instead we need to check
if it's set and not equal to our RODC number, allowing RODCs to trust
the full DCs and itself, but not other RODCs.
Andrew Bartlett
This remains an abuse, because it relies on setting into the krb5_principal
structure, but at least it causes less trouble for the server.
Andrew Bartlett
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Mon Nov 15 00:13:59 UTC 2010 on sn-devel-104
when we are an RODC and we get a request for a principal that we don't
have the right secrets for, we need to proxy the request to a
writeable DC. This happens for both TCP and UDP requests, for both
krb5 and kpasswd
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
Autobuild-User: Andrew Tridgell <tridge@samba.org>
Autobuild-Date: Fri Nov 12 08:03:20 UTC 2010 on sn-devel-104
This means that when we are an RODC, and an account does not have the
password attributes, we can now indicate to the kdc code that it
should forward the request to a real DC.
(The proxy code itself is not in this commit).
Andrew Bartlett
The new waf-based build system now has all the same functionality, and
the old build system has been broken for quite some time.
Autobuild-User: Jelmer Vernooij <jelmer@samba.org>
Autobuild-Date: Sun Oct 31 02:01:44 UTC 2010 on sn-devel-104
this also removes the event_context parameter from process model
initialisation. It isn't needed, and is confusing when a process model
init can be called from more than one place, possibly with different
event contexts.
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
This reverts commit 8a2ce5c47c.
Jelmer pointed out that these are also in use by other LDB databases - not only
SAMDB ones.
Autobuild-User: Matthias Dieter Wallnöfer <mdw@samba.org>
Autobuild-Date: Sun Oct 17 13:37:16 UTC 2010 on sn-devel-104
They're only in use by SAMDB code.
Autobuild-User: Matthias Dieter Wallnöfer <mdw@samba.org>
Autobuild-Date: Sun Oct 17 09:40:13 UTC 2010 on sn-devel-104
By setting the event context to use for this operation (only) onto
the krb5_context just before we call that operation, we can try
and emulate the specification of an event context to the actual send_to_kdc()
This eliminates the specification of an event context to many other
cli_credentials calls, and the last use of event_context_find()
Special care is taken to restore the event context in the event of
nesting in the send_to_kdc function.
Andrew Bartlett
We should avoid using Kerberos or any other recursive auth mechanism
in ldb backends, but denying Kerberos here won't be enough, so
remove the special case. (Typcially we bind using a different password
space and DIGEST-MD5 or NTLM).
Andrew Bartlett
This includes rewriting the PAC if the original krbtgt isn't to be
trusted, and reading different entries from the DB for the krbtgt
depending on the krbtgt number.
Andrew Bartlett
This means we just set up the system_session etc in one place
and don't diverge between the MIT and Heimdal plugins.
We also now determine if we are an RODC and store some details
that we will need later.
Andrew Bartlett
This makes the structure much more like NT_USER_TOKEN in the source3/
code. (The remaining changes are that privilages still need to be merged)
Andrew Bartlett
Now it's finally possible that the user can change his password with a DSDB
connection using his credentials.
NOTICE: I had to extract the old password from the SAMDB since I was unable to
find it somewhere else (authinfo for example).
this converts all callers that use the Samba4 loadparm lp_ calling
convention to use the lpcfg_ prefix.
Signed-off-by: Andrew Bartlett <abartlet@samba.org>
This changes the calculation to apply the allowed enc types to all
uses of the key (no point allowing a weak kinit to a key the server
wanted strongly protected). It also ensures that all the non-DES keys
are available on the krbtgt in particular, even as it does not have a
msds-SupportedEncryptionTypes attributes.
Andrew Bartlett
"heimdal/waf: Initial work on making it possible to use the system"
as the hdb_check_s4u2self function handling is incorrect.
This reverts commit b099631f42.
Adapt the two functions for the restructured "password_hash" module. This
means that basically all checks are now performed in the mentioned module.
An exception consists in the SAMR password change calls since they need very
precise NTSTATUS return codes on wrong constraints ("samr_password.c") file
Apparently AD supports setting a password on a servicePrincipalName,
not just a user principal name. This should fix (part of) the join of
OpenSolaris's internal CIFS server to Samba4 as reported by Bug #7273
Andrew Bartlett
Keep all heimdal related plugin code within hdb_samba4.c
Move interfaces needed by multiple plugins in db-glue.c
Move sequence context in main db context so that we do
not depend on db->hdb_dbc in the common code.
Remove unnecessary paremeters from function prototypes
This allows to use a common structure not tied to hdb_samba4
Also allows to avoid many casts within hdb_samba4 functions
This is the first step to abstract samba kdc databse functions
so they can be used by the MIT forthcoming plugin.
The change to use a common system_session broke replication as the KDC
forces CRED_DONT_USE_KERBEROS on session->credentials, which is shared
with other parts of the system.
This should be fixed once we confirm whether the ldap backend actually
relies on CRED_DONT_USE_KERBEROS
This patch adds a system_session cache, preventing us from having to
recreate it on every ldb open, and allowing us to detect when the same
session is being used in ldb_wrap
For KERBEROS applications the realm should be upcase (function "lp_realm") but
for DNS ones it should be used lowcase (function "lp_dnsdomain"). This patch
implements the use of both in the right way.
This code eventually calls into mk_priv in the Heimdal code, and if
the client is behind NAT, or somehow has an odd idea about it's own
network addresses, it will fail to accept this packet if we set an
address. It seems easiser not to. (Found by testing with NetAPP at
plugfest)
Andrew Bartlett
This extends the PKINIT code in Heimdal to ask the HDB layer if the
User Principal Name name in the certificate is an alias (perhaps just
by case change) of the name given in the AS-REQ. (This was a TODO in
the Heimdal KDC)
The testsuite is extended to test this behaviour, and the other PKINIT
certficate (using the standard method to specify a principal name in a
certificate) is updated to use a Administrator (not administrator).
(This fixes the kinit test).
Andrew Bartlett
It is much easier to do decryption with wireshark when the keytab is
available for every host in the domain. Running 'net export keytab
<keytab name>' will export the current (as pointed to by the supplied
smb.conf) local Samba4 doamin.
(This uses Heimdal's 'hdb' keytab and then the existing hdb-samba4,
and so has a good chance of keeping working in the long term).
Andrew Bartlett
It is much easier to do decryption with wireshark when the keytab is
available for every host in the domain. Running 'net export keytab
<keytab name>' will export the current (as pointed to by the supplied
smb.conf) local Samba4 doamin.
(This uses Heimdal's 'hdb' keytab and then the existing hdb-samba4,
and so has a good chance of keeping working in the long term).
Andrew Bartlett
This overloads the 'name' part of the keytab name to supply a context
pointer, and so avoids 3 global variables!
To do this, we had to stop putting the entry for kpasswd into the
secrets.ldb. (I don't consider this a big loss, and any entry left
there by an upgrade will be harmless).
Andrew Bartlett
This removes the last use of the prefix hdb_ldb and makes it clear
that we pass in 3 global variables to get state information into
hdb_samba4 when used as a keytab. (And that they belong to
hdb_samba4, not to the KDC)
Andrew Bartlett
To do this properly, we must use the PAC, but for now this is enough
to check that we are delegating to another name on the same host
(which must be safe).
(Windows 7 does this a lot, also noted in bug 6273)
Andrew Bartlett
Each attribute we request from LDB comes with a small cost, so don't
lookup any more than we must for the (very) frequent krbtgt lookup
case. Similarly, we don't need to build a PAC for a server (as a
target), so don't ask for the PAC attributes here either.
Andrew Bartlett
The previous code only allowed an KRB5_NT_ENTERPRISE name (an e-mail
list user principal name) in an AS-REQ. Evidence from the wild
(Win2k8 reportadely) indicates that this is instead valid for all
types of requests.
While this is now handled in heimdal/kdc/misc.c, a flag is now defined
in Heimdal's hdb so that we can take over this handling in future (once we start
using a system Heimdal, and if we find out there is more to be done
here).
Andrew Bartlett
The function LDB_lookup_principal() has been eliminated, and it's
contents spread back to it's callers. Removing the abstraction makes
the code clearer.
Also ensure we never pass unescaped user input to a LDB search
function.
Andrew Bartlett
This requires a rework on Heimdal's windc plugin layer, as we want
full control over what tickets Heimdal will issue. (In particular, in
case our requirements become more complex in future).
The original problem was that Heimdal's check would permit the ticket,
but Samba would then deny it, not knowing it was for kadmin/changepw
Also (in hdb-samba4) be a bit more careful on what entries we will
make the 'change_pw' service mark that this depends on.
Andrew Bartlett
Also including the supporting changes required to pass make test
A number of heimdal functions and constants have changed since we last
imported a tree (for the better, but inconvenient for us).
Andrew Bartlett
Logs showed that every SAM authentication was causing a non-indexed
ldb search for member=XXX. This was previously indexed in Samba4, but
since we switched to using the indexes from the full AD schema it now
isn't.
The fix is to use the extended DN operations to allow us to ask the
server for the memberOf attribute instead, with with the SIDs attached
to the result. This also means one less search on every
authentication.
The patch is made more complex by the fact that some common routines
use the result of these user searches, so we had to update all
searches that uses user_attrs and those common routines to make sure
they all returned a ldb_message with a memberOf filled in and the SIDs
attached.
It turns out (seen in MS-SAMR 3.1.1.7.1 for example) that the primary
way the krbtgt account is recognised as special is that RID. This
should fix issues such as 'password expired' on the kpasswd service.
Andrew Bartlett
A single AD server can only host a single domain, so don't stuff about
with looking up our crossRef record in the cn=Partitions container.
We instead trust that lp_realm() and lp_workgroup() works correctly.
Andrew Bartlett
When starting GENSEC on the server, the auth subsystem context must be
passed in, which now includes function pointers to the key elements.
This should (when the other dependencies are fixed up) allow GENSEC to
exist as a client or server library without bundling in too much of
our server code.
Andrew Bartlett
This uses a virtual attribute 'clearTextPassword' (name chosen to
match references in MS-SAMR) that contains the length-limited blob
containing an allegidly UTF16 password. This ensures we do no
validation or filtering of the password before we get a chance to MD4
it. We can then do the required munging into UTF8, and in future
implement the rules Microsoft has provided us with for invalid inputs.
All layers in the process now deal with the strings as length-limited
inputs, incluing the krb5 string2key calls.
This commit also includes a small change to samdb_result_passwords()
to ensure that LM passwords are not returned to the application logic
if LM authentication is disabled.
The objectClass module has been modified to allow the
clearTextPassword attribute to pass down the stack.
Andrew Bartlett
This avoids one more custom patch to the Heimdal code, and provides a
more standard way to produce hdb plugins in future.
I've renamed from hdb_ldb to hdb_samba4 as it really is not generic
ldb.
Andrew Bartlett
This is implemented by means of a message to the KDC, to avoid having
to link most of the KDC into netlogon.
Andrew Bartlett
(This used to be commit 82fcd7941f)
This uses Heimdal's PAC parsing code in the:
- LOCAL-PAC test
- gensec_gssapi server
- KDC (where is was already used, the support code refactored from here)
In addition, the service and KDC checksums are recorded in the struct
auth_serversupplied_info, allowing them to be extracted for validation
across NETLOGON.
Andrew Bartlett
(This used to be commit 418b440a7b)
At this stage, only arcfour-hmac-md5 trusts are used, and all trusts
are presumed bi-directional. Much more work still to be done.
Andrew Bartlett
(This used to be commit 3e9f5c2816)
