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.
> Here's the problem I hit:
>
> getgrnam("foo") -> nscd -> NSS -> winbindd ->
> winbindd_passdb.c:nam_to_sid() -> lookup_global_sam_name() ->
> getgrnam("foo") -> nscd -> ....
>
> This is in the SAMBA_3_0 specifically but in theory could happen
> SAMBA_3_0_25 (or 26) for an unknown group.
>
> The attached patch passes down enough state for the
> name_to_sid() call to be able to determine the originating
> winbindd cmd that came into the parent. So we can avoid
> making more NSS calls if the original call came in trough NSS
> so we don't deadlock ? But you should still service
> lookupname() calls which are needed for example when
> doing the token access checks for a "valid groups" from
> smb.conf.
>
> I've got this in testing now. The problem has shown up with the
> DsProvider on OS X and with nscd on SOlaris and Linux.
* Rely on the fact that name2sid will work for any name
in a trusted domain will work against our primary domain
(even in the absense of an incoming trust path)
* Only logons will reliably work and the idmap backend
is responsible for being able to manage id's without contacting
the trusted domain
* "getent passwd" and "getent group" for trusted users and groups
will work but we cannot get the group membership of a user in any
fashion without the user first logging on (via NTLM or krb5)
and the netsamlogon_cache being updated.
information return from our DC in the DsEnumerateDomainTrusts()
call. If the fails, we callback ot the older
connect-to-the-remote-domain method.
Note that this means we can only reliably expect the native_mode
flag to be set for our own domain as this information in not
available outside our primary domain from the trusted information.
This is ok as we only really need the flag when trying to
determine to enumerate domain local groups via RPC.
Use the AD flag rather than the native_mode flag when using
ldap to obtain the seq_num for a domain.
* Fix getgroups() call called using a normalized name
* Fix some more name mappings that could cause for example
a user to be unable to unlock the screen as the username
would not match in the PAM authenticate call.
we never mix malloc and talloc'ed contexts in the
add_XX_to_array() and add_XX_to_array_unique()
calls. Ensure that these calls always return
False on out of memory, True otherwise and always
check them. Ensure that the relevent parts of
the conn struct and the nt_user_tokens are
TALLOC_DESTROYED not SAFE_FREE'd.
James - this should fix your crash bug in both
branches.
Jeremy.
the moment) but winbindd isn't run in the build farm
so hopefully won't break anything too badly - I don't
want to lose this.
If winbindd starts offline then it falls back to using
MS-RPC backend. On going online it needs to reset the
backend and try and go to using the AD backend code if
possible, as the MS-RPC sequence number fetch just returns
1 as the sequence number if run against an AD DC.
In addition, the winbindd async child may end up
with the AD backend whilst the main winbindd - which
still contacts the DC for some non-async calls, is
left using MS-RPC. This can cause some trouble (as
you can imagine :-).
Attempt to ensure both main winbindd and async children
us AD backends on going online.
Jeremy.
* autogenerate lsa ndr code
* rename 'enum SID_NAME_USE' to 'enum lsa_SidType'
* merge a log more security descriptor functions from
gen_ndr/ndr_security.c in SAMBA_4_0
The most embarassing thing is the "#define strlen_m strlen"
We need a real implementation in SAMBA_3_0 which I'll work on
after this code is in.
Instead of trying to do this in the winbindd_cache
entries, add a timed even handler to probe every
5 mins when disconnected.
Fix events to run all pending events, rather than
only one.
Jeremy.
return NTSTATUS.
If we want to differentiate different name resolution problems we might want
to introduce yet another error class for Samba-internal errors. Things like no
route to host to the WINS server, a DNS server explicitly said host not found
etc might be worth passing up.
Because we can not stash everything into the existing NT_STATUS codes, what
about a Samba-specific error class like NT_STATUS_DOS and NT_STATUS_LDAP?
Volker
We may only feed rpc_useraliases with chunks of 1024 entries. This is
important as the token generation otherwise fails when a user is member
of more then 1024 groups.
Volker, please check.
Guenther
around failed query_user calls. This fixes
logons to a member of a Samba domain as a user from a
trusted AD domain.
As per comments on samba-technical, I still need to add
(a) cache the PAC info as werll as NTLM net_user_info_3
(b) expire the cache when the SMB session goes away
Both Jeremy and Guenther have signed off on the idea.
homedirectory and the loginshell from Active Directory's "Services for Unix".
Enable it with:
winbind sfu support = yes
User-Accounts without SFU-Unix-Attributes will be assigned template-based
Shells and Homedirs as before.
Note that it doesn't matter which version of Services for Unix you use (2.0,
2.2, 3.0 or 3.5). Samba should detect the correct attributes (msSFULoginShell,
msSFU30LoginShell, etc.) automatically.
If you also want to share the same uid/gid-space as SFU then also use PADL's
ad-idmap-Plugin:
idmap backend = ad
When using the idmap-plugin only those accounts will appear in Name Service
Switch that have those UNIX-attributes which avoids potential uid/gid-space
clashes between SFU-ids and automatically assigned idmap-ids.
Guenther
1. using smbc_getxattr() et al, one may now request all access control
entities in the ACL without getting all other NT attributes.
2. added the ability to exclude specified attributes from the result set
provided by smbc_getxattr() et al, when requesting all attributes,
all NT attributes, or all DOS attributes.
3. eliminated all compiler warnings, including when --enable-developer
compiler flags are in use. removed -Wcast-qual flag from list, as that
is specifically to force warnings in the case of casting away qualifiers.
Note: In the process of eliminating compiler warnings, a few nasties were
discovered. In the file libads/sasl.c, PRIVATE kerberos interfaces
are being used; and in libsmb/clikrb5.c, both PRIAVE and DEPRECATED
kerberos interfaces are being used. Someone who knows kerberos
should look at these and determine if there is an alternate method
of accomplishing the task.
whole of samba comiles warning-free with the default compiler flags.
Temporarily defined -Wall to locate other potential problems. Found an
unused static function (#ifdefed out rather than deleted, in case it's
needed for something in progress).
There are also a number of uses of undeclared functions, mostly krb5_*.
Files with these problems need to have appropriate header files included,
but they are not fixed in this update.
oplock_linux.c.c has undefined functions capget() and capset(), which need
to have "#undef _POSIX_SOURCE" specified before including <sys/capability.h>,
but that could potentially have other side effects, so that remains uncorrected
as well.
The flag -Wall should be added permanently to CFLAGS, and all warnings then
generated should be eliminated.
based on samba4-idl.
This saves us an enormous amount of totally unnecessary ldap-traffic
when several hundreds of winbind-daemons query a Samba3 DC just to get
the fake SAM-sequence-number (time(NULL)) by enumerating all users, all
groups and all aliases when query-dom-info level 2 is used.
Note that we apparently never get the sequence number right (we parse a
uint32, although it's a uint64, at least in samba4 idl). For the time
being, I would propose to stay with that behaviour.
Guenther
