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.
Handle the case where a NT4 DC does not fill in the acct_flags in
the samlogon reply info3. Yes, in 2021, there are still admins
arround with real NT4 DCs.
NT4 DCs reject authentication with workstation accounts with
NT_STATUS_NOLOGON_WORKSTATION_TRUST_ACCOUNT, even if
MSV1_0_ALLOW_WORKSTATION_TRUST_ACCOUNT is specified.
We no longer call dcerpc_samr_QueryUserInfo(level=16)
to get the acct_flags, as we only ever got
ACB_NORMAL back (maybe with ACB_PWNOEXP in addition),
which is easy to calculate on our own.
This was removed in commit (for 4.15.0rc1):
commit 73528f26eea24033a7093e5591b8f89ad2b8644e
Author: Ralph Boehme <slow@samba.org>
AuthorDate: Mon Jan 11 14:59:46 2021 +0100
Commit: Jeremy Allison <jra@samba.org>
CommitDate: Thu Jan 21 22:56:20 2021 +0000
winbind: remove legacy flags fallback
Some very old NT4 DCs might have not returned the account flags filled in. This
shouldn't be a problem anymore. Additionally, on a typical domain member server,
this request is (and can only be) send to the primary domain, so this will not
work with accounts from trusted domains.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Jan 21 22:56:20 UTC 2021 on sn-devel-184
It means one more caller of the problematic cm_connect_sam()
function is removed! SAMR connections may not be allowed for
machine accounts with modern AD DCs.
For network logons NT4 DCs also skip the
account_name, so we have to fallback to the
one given by the client. We have code to cope
with that deeply hidden inside of netsamlogon_cache_store().
Up to Samba 4.7 netsamlogon_cache_store() operated on the
info3 structure that was passed to the caller of winbind_dual_SamLogon()
and pass propagated up to auth_winbind in smbd.
But for Samba 4.8 the following commit:
commit f153c95176b7759e10996b24b66d9917945372ed
Author: Ralph Boehme <slow@samba.org>
Date: Mon Dec 11 16:25:35 2017 +0100
winbindd: let winbind_dual_SamLogon return validation
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
actually changed the situation and only a temporary info3 structure
was passed into netsamlogon_cache_store(), which means
account_name was NULL and get propagated as "" into auth_winbind
in smbd, where getpwnam() is no longer possible and every
smb access gets NT_STATUS_LOGON_FAILURE.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14772
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Aug 3 11:10:27 UTC 2021 on sn-devel-184
The new code uses PING_DC to tell the child to try to go online.
Pair-Programmed-With: Andreas Schneider <asn@samba.org>
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Isaac Boukris <iboukris@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Autobuild-User(master): Andreas Schneider <asn@cryptomilk.org>
Autobuild-Date(master): Thu Apr 29 15:49:16 UTC 2021 on sn-devel-184
The sequence number is used to detect if the cache is still valid. It
expires when the `winbind cache time` is over. After that time we want
to fetch new information from a DC to make sure we are up to date.
If a DC goes down and we recreate the connection, we want to expire the
caches sooner. So we reset the sequence number and the next call should
refill the caches.
Using the current time as the sequence number is more reliable, as the
sequence number of two DCs could in theory be equal. All we have to do
is to make sure we reset it after we reconnect to a DC.
Previously the sequence number check was based on the AD database change
sequence number. Now this is based on a current time value which gets
reset after a successful (re)connect.
Pair-Programmed-With: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Ralph Boehme <slow@samba.org>
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14695
Signed-off-by: Andreas Schneider <asn@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Thu Apr 29 04:48:37 UTC 2021 on sn-devel-184
Those calls were only used in winbindd_samr which now does direct and
simpler samr calls.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Same argument as with previous patches: We don't need fancy lsa
routing and samr is less prone to deadlock back into winbind
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Same argument as with name_to_sid: We don't need the lsa lookup
routing, and samr is less prone to deadlocking.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
After the "Unix Users/Groups" and wkn names have been taken care of,
all that remains here is our domain (BUILTIN or workgroup). We don't
need any of the fancy routing in lsa_lookupnames, and samr_LookupNames
is a lot less prone to deadlocks back into winbind.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
"Unix Users" and "Unix Groups" can recurse into nsswitch and thus into
winbind. In the binding process, we have winbindd_off(), but if we
pass the lookupNames request to a forked lsad, lsad does not
necessarily have that setting. So lsad might turn back to winbind,
which could lead to a deadlock. Handle the nsswitch lookups in
winbind.
While there, also do the simple wellknown names and the "DOMAIN\" type
3 lookups directly in winbind.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
If you try to allocate a GID before winbind is fully set up,
idmap_child_handle() is still NULL. Add the required
wb_parent_idmap_setup_send()/recv() to allocate_gid().
Bug: https://bugzilla.samba.org/show_bug.cgi?id=14678
RN: Fix a crash in winbind when allocate-gid is called early
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
is_myname() looks at lp_* directly, nmbd maintains its own list: We don't
need the baroque loadparm handler anymore.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
This structure just grew from 3 to 6 pointers, avoid making a copy of
this. All callers of dcesrv_init_context() have this as a static
struct in the C object, so a pointer to that won't change.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Not used right now, but we should never have callbacks without a
"private_data" pointer. Some of the callbacks could even today benefit
from this.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
The call to winbind_lookup_name already wrote the result in the id_map
array. The later check for the type detected a mismatch, but that did
not remove the SID from the result struct.
Change this by first assigning the SID to a temporary variable and only
write it to the id_map array after the type checks.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14663
Signed-off-by: Christof Schmitt <cs@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Thu Mar 11 08:38:41 UTC 2021 on sn-devel-184
The call to winbind_lookup_name already wrote the result in the id_map
array. The later check for the type detected a mismatch, but that did
not remove the SID from the result struct.
Change this by first assigning the SID to a temporary variable and only
write it to the id_map array after the type checks.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14663
Signed-off-by: Christof Schmitt <cs@samba.org>
Only consider a mapping to be valid when the idmap module reports
ID_MAPPED. Otherwise return the null SID.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14663
Signed-off-by: Christof Schmitt <cs@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
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): Tue Mar 9 23:45:51 UTC 2021 on sn-devel-184
Before merging the s4 and s3 RPC servers the make_internal_rpc_pipe_p()
function did not fail when the requested interface was not registered in
the calling process because it did not check the return value of
rpc_srv_get_pipe_cmds(). If the interface was not registed, the pointer
to the interface functions was NULL and later, when dispatching a call,
rpcint_dispatch() returned NT_STATUS_RPC_PROCNUM_OUT_OF_RANGE in this
case.
After merging the RPC servers, the rpc_pipe_open_internal() function
will return NT_STATUS_RPC_INTERFACE_NOT_FOUND if the interface is not
registered in the calling process. This causes a noisy error message in
winbind when it tries to open the dssetup pipe to the primary domain and
it is not an AD domain.
The callers of wb_open_internal_pipe() when connecting to the domain
already logs the error at level greather or equal to five. This commit
moves the dupplicated and noisy error message at level zero from
wb_open_internal_pipe() to its callers outside winbindd_cm.c.
This error can be seen in winbindd logs of ad_member and nt4_member test
environments.
[2021/03/01 16:49:38.486004, 0, pid=12456] ../../source3/winbindd/winbindd_cm.c:1893(wb_open_internal_pipe)
open_internal_pipe: Could not connect to dssetup pipe: NT_STATUS_RPC_INTERFACE_NOT_FOUND
Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
When samba runs as ADDC only the main 'samba' daemon have to notify
its status to systemd because our systemd unit files contains implied
NotifyAccess=main since commit d1740fb3d5a72cb49e30b330bb0b01e7ef3e09cc.
This commit adds a function to disable the systemd notification in the
smbd and winbinddd child processes started by the main 'samba' daemon in
AD DC mode to avoid warnings like:
systemd[1]: samba-ad-dc.service: Got notification message from PID 26194,
but reception only permitted for main PID 26187
systemd[1]: samba-ad-dc.service: Got notification message from PID 26222,
but reception only permitted for main PID 26187
$ pstree -p
...
├─samba(26187)─┬─tfork(26189)(26188)───s3fs[master](26189)───tfork(26194)(26193)───smbd(26194)─┬─cleanupd(+
│ │ ├─lpqd(2623+
│ │ └─smbd-noti+
│ ├─tfork(26191)(26190)───rpc[master](26191)─┬─tfork(26198)(26195)───rpc(0)(26198)
│ │ ├─tfork(26200)(26199)───rpc(1)(26200)
│ │ ├─tfork(26206)(26201)───rpc(2)(26206)
│ │ └─tfork(26212)(26207)───rpc(3)(26212)
│ ├─tfork(26196)(26192)───nbt[master](26196)
│ ├─tfork(26202)(26197)───wrepl[master](26202)
│ ├─tfork(26204)(26203)───ldap[master](26204)─┬─tfork(26242)(26241)───ldap(0)(26242)
│ │ ├─tfork(26244)(26243)───ldap(1)(26244)
│ │ ├─tfork(26246)(26245)───ldap(2)(26246)
│ │ └─tfork(26248)(26247)───ldap(3)(26248)
│ ├─tfork(26208)(26205)───cldap[master](26208)
│ ├─tfork(26210)(26209)───kdc[master](26210)───tfork(26218)(26215)───krb5kdc(26218)
│ ├─tfork(26213)(26211)───drepl[master](26213)
│ ├─tfork(26216)(26214)───winbindd[master(26216)───tfork(26222)(26219)───winbindd(26222)───wi+
│ ├─tfork(26220)(26217)───ntp_signd[maste(26220)
│ ├─tfork(26223)(26221)───kcc[master](26223)
│ ├─tfork(26225)(26224)───dnsupdate[maste(26225)
│ └─tfork(26227)(26226)───dns[master](26227)
Signed-off-by: Samuel Cabrero <scabrero@samba.org>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>
All callers manually dissected the syntax id for this API.
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Samuel Cabrero <scabrero@samba.org>
idmap_hash used to bounce back the requested type,
which was ID_TYPE_UID, ID_TYPE_GID or ID_TYPE_NOT_SPECIFIED
before as the winbindd parent always used a lookupsids.
When the lookupsids failed because of an unknown domain,
the idmap child weren't requested at all and the caller
sees ID_TYPE_NOT_SPECIFIED.
This module should have supported ID_TYPE_BOTH since
samba-4.1.0, similar to idmap_rid and idmap_autorid.
Now that the winbindd parent will pass ID_TYPE_BOTH in order to
indicate that the domain exists, it's better to always return
ID_TYPE_BOTH instead of a random mix of ID_TYPE_UID, ID_TYPE_GID
or ID_TYPE_BOTH. In order to request a type_hint it will return
ID_REQUIRE_TYPE for ID_TYPE_NOT_SPECIFIED, which means that
the parent at least assures that the domain sid exists.
And the caller still gets ID_TYPE_NOT_SPECIFIED if the
domain doesn't exist.
BUG: https://bugzilla.samba.org/show_bug.cgi?id=14539
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Fri Jan 22 11:32:46 UTC 2021 on sn-devel-184
Some very old NT4 DCs might have not returned the account flags filled in. This
shouldn't be a problem anymore. Additionally, on a typical domain member server,
this request is (and can only be) send to the primary domain, so this will not
work with accounts from trusted domains.
Signed-off-by: Ralph Boehme <slow@samba.org>
Reviewed-by: Jeremy Allison <jra@samba.org>
Autobuild-User(master): Jeremy Allison <jra@samba.org>
Autobuild-Date(master): Thu Jan 21 22:56:20 UTC 2021 on sn-devel-184
