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 means "xcopy /o" has a chance of working with ACLs that contain
ACEs that use SIDs that the Samba server has no knowledge of.
It's a bit hackish, Tim, can you look at my uid.c changes?
(This used to be commit fe2db31485)
This patch fixes the segfaults I introduced in the previous conneciton caching
patch. It cleans up the connection cache a *lot* - in particular it adds
significant robustness to the operation.
If a the DC goes down, we no longer fail the next operation - the code checks
if the connection died during one of its own operations on the socket, and
restarts the conneciton as required.
There is still a memory leak in here somewhere - but this code also cleans up a
number of these.
Also added is the abilty to sepecify the domain of the 'get around restrict anonymous'
user that winbind uses.
Andrew Bartlett
(This used to be commit 92cbefdf27)
Add a connection cache to the netlogon pipe. This makes a *massive* difference
to the time-per-auth. Also fix up *some* of the memory leaks in other
connection caches.
Add some debugging messages for the is_connected() code. I'm thinking we
should get a client implementation of SMBecho and call it here - as it would
allow us to always know the DC is around before we start.
Down the debug level for some of the pam_winbind code - I'll probably down it
further when I'm finished debugging.
Andrew Bartlett
(This used to be commit 49d3e47666)
This adds code to do generic PAM -> NTSTATUS and NTSTATUS -> PAM error
conversions, and uses them to make the error handling in pam_winbind sane.
In particular, pam_winbind now uses PAM error codes, not silly '-1, -2 ...'
stuff, and logs the NTSTATUS error that winbind now sends over the pipe.
Added code to wbinfo to display these - makes a big difference in debugging
winbindd.
The main change here is the code to allow pam_winbind password changing to
correctly stack - This code ripped from pam_unix, and the copyright attached.
(Same as for all pam modules, including pam_winbind)
Andrew Bartlett
(This used to be commit dc1a72f896)
processing work correctly in winbindd. This is a really good patch
that gives full select semantics to the Samba modified select.
Jeremy.
(This used to be commit 3af16ade17)
<a.bokovoy@sam-solutions.net>.
This patch is designed to remove the 'special cases' required for this support.
In particular this now kills off winbind_initgroups, as it appears no longer to
be required.
Andrew Bartlett
(This used to be commit f1d8d50976)
winbind default domains, particulary now I understand whats going on a lot
better. This ensures that the RPC client code does as little 'magic' as
possible - this is up to the application/user. (Where - for to name->sid code
- it was all along). This leaves the change that allows the sid->name code to
return domains and usernames in seperate paramaters.
Andrew Bartlett
(This used to be commit 5dfba2cf53)
the sid->uid and uid->sid conversions.
Remove some duplicate arguments from these funcitons, and update the
request/response structures for this and the 'winbind domain name' feature.
As such 'winbindd_lookup_name' now takes both a domain and username.
(This used to be commit ce1b4d4c30)
the "password server" smb.conf parameter when choosing a DC to connect to.
Due to the origin of the code in cm_get_dc_name() it wouldn't try
additional DCs if the first DC didn't work. This would wedge winbindd if you
had "password server = foo1, foo2" and foo1 was down.
(This used to be commit fc7ed1b4a8)
smbd, and also makes it much cleaner inside winbindd.
It is mostly my code, with a few changes and testing performed by Alexander
Bokovoy <a.bokovoy@sam-solutions.net>. ab has tested it in security=domain and
security=ads, but more testing is always appricatiated.
The idea is that we no longer cart around a 'domain\user' string, we keep them
seperate until the last moment - when we push that string into a pwent on onto
the socket.
This removes the need to be constantly parsing that string - the domain prefix
is almost always already provided, (only a couple of functions actually changed
arguments in all this).
Some consequential changes to the RPC client code, to stop it concatonating the
two strings (it now passes them both back as params).
I havn't changed the cache code, however the usernames will no longer have a
double domain prefix in the key string. The actual structures are unchanged
- but the meaning of 'username' in the 'rid' will have changed. (The cache is
invalidated at startup, so on-disk formats are not an issue here).
Andrew Bartlett
(This used to be commit e870f0e727)
<a.bokovoy@sam-solutions.net>.
The idea is the domain\username is rather harsh for unix systems - people don't
expect to have to FTP, SSH and (in particular) e-mail with a username like
that.
This 'corrects' that - but is not without its own problems.
As you can see from the changes to files like username.c and wb_client.c (smbd's
winbind client code) a lot of assumptions are made in a lot of places about
lp_winbind_seperator determining a users's status as a domain or local user.
The main change I will shortly be making is to investigate and kill off
winbind_initgroups() - as far as I know it was a workaround for an old bug in
winbind itself (and a bug in RH 5.2) and should no longer be relevent.
I am also going to move to using the 'winbind uid' and 'winbind gid' paramaters
to determine a user/groups's 'local' status, rather than the presence of the
seperator.
As such, this functionality is recommended for servers providing unix services,
but is currently less than optimal for windows clients.
(TODO: remove all references to lp_winbind_seperator() and
lp_winbind_use_default_domain() from smbd)
Andrew Bartlett
(This used to be commit 07a21fcd23)
memory.
The winbind connection caching code isn't exactly a plesent beast, and there is
more work that needs to be done to nail this properly.
Andrew Bartlett
(This used to be commit dd40ce54b7)
to move this from being a static to matching its mate in lib/util_sock.c.
In any case, this should discorage anybody from using the 'wrong' version of
this function. (ie the one from TNG, which needs a bit more error checking
depending on use).
Andrew Bartlett
(This used to be commit e6a3a01f79)
This work was sponsored by Optifacio Software Services, Inc.
Andrew Bartlett
(various e-mails announcements merged into some form of commit message below:)
This patch which adds basics of universal groups support
into Samba 3. Currently, only Winbind with RPC calls supports this, ADS
support requires additional (possibly huge) work on KRB5 PAC. However,
basic infrastructure is here.
This patch adds:
1. Storing of universal groups for particular user logged into Samba
software (smbd/ two winbind-pam methods) into netlogon_unigrp.tdb as array
of uint32 supplemental group rids keyed as DOMAIN_SID/USER_RID in tdb.
2. Fetching of unversal groups for given user rid and domain sid from
netlogon_unigrp.tdb.
Since this is used in both smbd and winbindd, main code is in
source/lib/netlogon_uingrp.c. Dependencies are added to AUTH_OBJ as
UNIGRP_OBJ and WINBINDD_OBJ as UNIGRP_OBJ.
This patch has had a few versions, the final version in particular:
Many thanks to Andrew Bartlett for critics and comments, and partly
rewritten code.
New:
- updated fetching code to changed byte order macros
- moved functions to proper namespace
- optimized memory usage by reusing caller's memory context
- enhanced code to more follow Samba coding rules
Todo:
- proper universal group expiration after timeout
(This used to be commit 80c2aefbe7)
the list received at startup or we get an out of date list. I thought
there might be some sequence number that is incremented when a trusted
domain is added or removed - perhaps there is but I just haven't found it
yet.
- Renamed get_domain_info() to init_domain_list()
- Made an accessor function to return the list of trusted domains rather
than using a global so we don't have to remember to put a magic init
function
- The getent state can not keep a pointer to a winbind_domain structure as
it may be freed if init_domain_list() is called again so we keep the
domain name instead
(This used to be commit 37216c649a)
swedish" test to client calls. This is putting a length field at the
start of a request so we can disconnect clients talking with an out of date
libnss_winbind.so rather than deadlock them.
Misc cleanups:
- made some int values uint32
- moved WINBIND_INTERFACE_VERSION to start of cmd list
(This used to be commit a4af65b9b9)
smb.conf to get it right.
While wb_client needs its lp_load() for samba dependency reasons, it now uses
the new method both to example and test the new code.
Also add an interface version function, and return the winbind's samba version
string.
In preperation for default domains, its now up to winbindd to reject plaintext
auths that don't have a seperator, but NTLM (CRAP) auths now have two feilds,
hence need parsing.
Andrew Bartlett
(This used to be commit 2bd2a092ee)
It adds a 'ping' request, just to check winbind is in fact alive
It also changes winbindd_pam_auth_crap to take usernames and domain seperatly.
(backward incompatible change, needs merge to 2.2, but this is not yet released
code, so no workarounds)
Finally, it adds some debugs and fixes a few memory leaks (uses talloc to do
it).
Andrew Bartlett
(This used to be commit 6df29bfe33)
when they are added or removed on the PDC.
- renamed GETPWNAM_FROM_{UID,USER} constants and functions to GETPW{NAM,UID}
- renamed GETGRNAM_FROM_{GID,GROUP} constants and functions to GETGR{NAM,GID}
- use SIGUSR2 in winbindd for debugging/logging instead of SIGUSR1 in
preparation for moving to smbcontrol type messages (not sure whether to
ditch this altogether or not)
- tidy debugging messages in top level winbind user and group routines
- convert talloc_init() to talloc_init_named()
- make enumerations of the domain list use the same local variable names
(This used to be commit eeb8af9c1a)
Also removed the dependency on auth_util.o, which makes things nicer.
Finally, this kills off the NECESSARY_BECAUSE_SAMBA_DEPENDENCIES_ARE_SO_BROKEN_OBJ
makefile variable - becouse Samba dependencies are starting to be sane again!
Andrew Bartlett
(This used to be commit 4609edcac3)
the method used for checking if a domain is a trusted domain is very
crude, we should really call a backend fn of some sort. For now I'm
using winbindd to do the dirty work.
(This used to be commit adf44a9bd0)
Added a --set-auth-user function to set a username and password that can be
used by winbindd when making connections to domain controllers. This is
necessary when restrictions have been placed on anonymous connections
either through the RestrictAnonymous registry setting, or the win2k Local
Security Policy -> Security Settings -> Local Policies -> Security Options
-> Additional restrictions for anonymous connections. (phew)
Two new keys are set in secrets.tdb: SECRETS/AUTH_USER and
SECRETS/AUTH_PASSWORD which hold the username and plaintext password of the
user to connect as.
To reset these values, run wbinfo --set-auth-user ""
(This used to be commit 507003522b)
I tried testing this by lowering the buffer size in
cli_samr_enum_dom_groups() but that didn't work - I think this needs
more looking into
(This used to be commit 34328e3031)
consistent backends (like ADS) always give correct primary group
info, so we can play cache tricks to speed things up a lot
inconsistent backends (like MSRPC) need to fetch stuff more often
(This used to be commit 217c39f232)
the backends
at startup, loop until we get the domain sid for our primary domain,
trying every 10 seconds. This makes winbindd handle a room-wide power
failure better
(This used to be commit 7c60ae5937)
this one looks like just another winbind backend, and has the
following properties:
- does -ve and +ve cacheing of all queries
- can be disabled with -n switch to winbindd
- stores all records packed, so even huge domains are not a problem
for a complete cache
- handles the server being down
- uses sequence numbers for all entries
This fixes a lot of problems with winbindd. Serving from cache is now
*very* fast.
(This used to be commit fddb4f4c04)
winbindd can do a kinit
this will be removed once we have code that gets a tgt
and puts it in a place where cyrus-sasl can see it
(This used to be commit 7d94f1b736)
I spent quite a while trying to work out how to make this call
via ldap and failed. I then found that MS servers seem use rpc
for sid_to_name, and it works even when in native mode, I ended
up just implementing it via rpc
(This used to be commit 789833b44e)
also created winbindd_rpc.c which contains the functions that
have been converted to the new structure. There will soon be
a winbindd_ads.c for the ldap backend
(This used to be commit e4ccc602ba)
This just splits off the dispinfo call behind a methods structure.
I'll split off a few more functions soon, then we will be ready for
LDAP replacement methods
(This used to be commit 0216b0fca1)
name_status_find() call here should look up a #1c name instead of #1d.
This fixes some bugs currently with BDC authentication in winbindd and in
smbd as you can't query the #1d name with the ip address of a BDC.
Who is Uncle Tom Cobbley anyway?
(This used to be commit 4215048f7b)
subystem.
The particular aim is to modularized the interface - so that we
can have arbitrary password back-ends.
This code adds one such back-end, a 'winbind' module to authenticate
against the winbind_auth_crap functionality. While fully-functional
this code is mainly useful as a demonstration, because we don't get
back the info3 as we would for direct ntdomain authentication.
This commit introduced the new 'auth methods' parameter, in the
spirit of the 'auth order' discussed on the lists. It is renamed
because not all the methods may be consulted, even if previous
methods fail - they may not have a suitable challenge for example.
Also, we have a 'local' authentication method, for old-style
'unix if plaintext, sam if encrypted' authentication and a
'guest' module to handle guest logins in a single place.
While this current design is not ideal, I feel that it does
provide a better infrastructure than the current design, and can
be built upon.
The following parameters have changed:
- use rhosts =
This has been replaced by the 'rhosts' authentication method,
and can be specified like 'auth methods = guest rhosts'
- hosts equiv =
This needs both this parameter and an 'auth methods' entry
to be effective. (auth methods = guest hostsequiv ....)
- plaintext to smbpasswd =
This is replaced by specifying 'sam' rather than 'local'
in the auth methods.
The security = parameter is unchanged, and now provides defaults
for the 'auth methods' parameter.
The available auth methods are:
guest
rhosts
hostsequiv
sam (passdb direct hash access)
unix (PAM, crypt() etc)
local (the combination of the above, based on encryption)
smbserver (old security=server)
ntdomain (old security=domain)
winbind (use winbind to cache DC connections)
Assistance in testing, or the production of new and interesting
authentication modules is always appreciated.
Andrew Bartlett
(This used to be commit 8d31eae52a)
winbindd_lookup_sid_by_name. Also if the lookup fails then clobber
the output parameters rather than leaving them looking potentially
valid.
Add doxygen.
(This used to be commit 61dba52a54)
in smbd/process.c where the timezone is reinitialised. Was replaced with
check for a static is_initialised boolean.
(This used to be commit 8fc772c9e5)
There was a bug in cli_lsa_lookup_name/lookup_sid where NT_STATUS_NONE_MAPPED was
being mapped to NT_STATUS_OK, and also the *wrong* number of entries mapped
was being returned. The correct field is mapped_count, *NOT* num_entries.
Jeremy.
(This used to be commit 9f8c644abc)
be easily added (a one liner) once we know the correct error codes returned
by a W2K DC.
All other winbindd calls should go through a similar transparent caching layer
(and will soon).
Jeremy.
(This used to be commit b16bb21d37)
Ignore the SIGUSR1 signal before we install a handler for it as glibc (?)
seems to just print out "User defined signal 1" and exit if no handler
is installed.
(This used to be commit 1212591095)
