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.
multiple event contexts
It is NEVER valid to free an event context that anybody else may have
a reference to, and never normally valid to have two 'live' at once.
We must instead call tevent_re_initialise() to wipe clean an existing
pointer.
Andrew Bartlett
This will fork off exactly one child to handle some task, ensuring
that if it dies or changes global state, that this does not change
everything.
Andrew Bartlett
The auth context was in the past only for NTLM authentication, but we
need a SAM, an event context and and loadparm context for calculating
the local groups too, so re-use that infrustructure we already have in
place.
However, to avoid problems where we may not have an auth_context (in
torture tests, for example), allow a simpler 'session_info' to be
generated, by passing this via an indirection in gensec and an
generate_session_info() function pointer in the struct auth_context.
In the smb_server (for old-style session setups) we need to change the
async context to a new 'struct sesssetup_context'. This allows us to
use the auth_context in processing the authentication reply .
Andrew Bartlett
We need to be able to give sensible error messages when a kerberos
calls fails. This propogates the kerberos error up the stack to the
caller.
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
This reverts commit 69d5cea2e5.
This commit causes issues with the RPC server, revert it until we find the
exact issue and possibly have a torture test to avoid it happening again.
Found playing with w2k8r2 and forest trusts.
After looking at the s4 side of the (s)channel :) I found out that it makes
more sense to simply make it use the tdb based code than redo the same changes
done to s3 to simplify the interface.
Ldb is slow, to the point it needs haks to pre-open the db to speed it up, yet
that does not solve the lookup speed, with ldb it is always going to be slower.
Looking through the history it is evident that the schannel database doesn't
really need greate expanadability. And lookups are always done with a single
Key. This seem a perfet fit for tdb while ldb looks unnecessarily complicated.
The schannel database is not really a persistent one. It can be discared during
an upgrade without causing any real issue. all it contains is temproary session
data.
We need to free the lp_ctx after we free the event context, otherwise
the teardown code in the notify backend dies when it tries to use the
iconv_convenience ptr
Fixes bug 7053
This task watches for changes in the list of DCs, and creates a bind9
formatted file that grants update permission to all DCs, plus to the
administration, and machines update for their own names.
Pair-Programmed-With: Andrew Bartlett <abartlet@samba.org>
A KCC is a 'Knowledge Consistency Checker', a fancy name for a daemon
that works out who will replicate with who in a AD domain. This
implements an extremely simple KCC task that just wants to replicate
with everyone :-)
Previously we relied on process groups and SIGTERM to ensure that
child tasks died in the standard process model when the parent task
died. This doesn't work when the server is run in interactive mode, as
in that case we don't call become_daemon() and don't get a separate
process group.
The fix is to have a pipe held open by the parent server process, and
inherited by child tasks. If the parent exits then the write side of
the pipe is implicitly closed, which causes an event in the child
tasks that causes them to exit
While testing the use of the standard process model with 'make test' I
found that testing was much slower (by several times) with the
standard model than with the single model. The primary problem was
that each SMB connection would open a new sam ldb context, and all of
those would reload the full AD schema.
The fix is to pre-open the SAM during server startup, before any child
processes are forked. This sets up the global schema context which is
inherited by all connections.
The standard model is still slower at make test than the single model,
but not by nearly as much. I am working on further reducing the gap.
For now implement just a very simple protocol
to negotiate the transport session info.
We may need to pass more info later,
e.g. client ip or name...
metze
The problem was that the parent smbd doesn't have
any event when the process model standard was in use.
Now we always add an fd event for stdin, but may not
ask for any events to trigger.
metze
(This used to be commit fa6c00734c)
The problem here was that with the packet code set to serialise, we
can have multiple packets 'processing' at once, and previously the
second packet (allowed because we are spining on an event context down
the stack) would clear the flag.
Andrew Bartlett
(This used to be commit 3378911124)
specific debug function.
By default do not debug, this is the most appropriate action for a library
as we cannot assume what stderr is use for in the main app.
The main app is responsible to set ev_debug_stderr if they so desire.
(This used to be commit e566a2f308)
NULL represents an anoymous connection, as that's
cheaper than always calling auth_anonymous_session_info(),
That means if the caller wants to use the transport layer
session_info he should call auth_anonymous_session_info()
to create the session info.
metze
(This used to be commit 3777d0307b)
Fix talloc_steal in both cases to steal connected socket (from accept)
onto it's private structure, rather than stealing the bound socket.
Remove termination code from the prefork modal, we want the process to
still stay around, to serve future clients.
Andrew Bartlett and David Disseldorp
(This used to be commit 07590d893e)
To use, run 'smbd -M prefork'
By default, only the smb service is preforked. 4 children are
created, and all listen for new connections. The Linux Kernel 'wake
one' behaviour should ensure that only one is given the oportunity to
accept. We need to look into the ideal number of worker children, as
well as load balancing behaviours.
To change, set:
prefork children : smb = 6
valid service names (smb in this example) match those in 'server services'.
Andrew Bartlett and David Disseldorp
(This used to be commit 35313c0aa3)
2007-09-29 More higher-level passing around of lp_ctx.
2007-09-29 Fix warning.
2007-09-29 Pass loadparm contexts on a higher level.
2007-09-29 Avoid using global loadparm context.
(This used to be commit 3468952e77)
been working on for at least half a year now. Contains the following
improvements:
* proper layering (finally!) for the registry library. Distinction is
now made between 'real' backends (local, remote, wine, etc) and
the low-level hive backends (regf, creg, ldb, ...) that are only used
by the local registry backend
* tests for all important hive and registry operations
* re-enable RPC-WINREG tests (still needs more work though, as
some return values aren't checked yet)
* write support for REGF files
* dir backend now supports setting/reading values, creating keys
* support for storing security descriptors
* remove CREG backend as it was incomplete, didn't match the data model
and wasn't used at all anyway
* support for parsing ADM files as used by the policy editor (see lib/policy)
* support for parsing PREG files (format used by .POL files)
* new streaming interface for registry diffs (improves speed and memory usage
for regdiff/regpatch significantly)
... and fixes a large number of bugs in the registry code
(This used to be commit 7a1eec6358)
