summaryrefslogtreecommitdiff
path: root/source4
AgeCommit message (Collapse)AuthorFilesLines
2009-10-25s4-samdb: reduce the number of samdb opens at startupAndrew Tridgell6-44/+13
Using common parameters means that the ldb_wrap code can return a reference rather than a new database
2009-10-25s4-ldb: ensure DNs pass validity tests in indexingAndrew Tridgell1-0/+4
2009-10-25s4-ldb: fixed string length handling on index recordsAndrew Tridgell1-2/+2
2009-10-25s4-dsdb: ensure that new partitions inherit any transactionAndrew Tridgell3-11/+26
2009-10-25s4-ldb: don't allow modifies outside a transaction.Andrew Tridgell1-0/+8
2009-10-25s4-ldb: fixed re-index during a complex transactionAndrew Tridgell1-10/+31
We may have modified index objects in the in-memory index tdb
2009-10-25s4-python: fixed annoyance where control-C doesn't kill our python scriptsAndrew Tridgell1-0/+10
We want our scripts to die immediately when a user hits control-C. Otherwise we not only annoy the hell out of the user, we also risk db corruption as the control-C could get delivered as an exception which gets mis-interpreted (eg. as a missing db object). We use transactions for all our databases, so the right thing to do in all our command line tools is to die immediately.
2009-10-24s4:dcesrv_samr: always use mem_ctx as initial parent for samr_*_stateStefan Metzmacher1-7/+7
We always steal the state to the policy handle on success, but untill then keep it on the short term context. metze
2009-10-24s4:loadparm: don't leak the names of all shares in each lp_service() callStefan Metzmacher1-1/+4
metze
2009-10-24s4:dsdb/partition_init: don't leak a talloc_new() in case we have no data yetStefan Metzmacher1-2/+5
metze
2009-10-24s4:gensec/schannel: remove unused talloc_reference() in schannel_update()Stefan Metzmacher1-1/+1
We never expose creds to the caller in schannel_update(). metze
2009-10-23ldb python bindungs - better use the "enum ldb_scope" for the search scope ↵Matthias Dieter Wallnöfer1-2/+3
rather than "int"
2009-10-23s4-python: we need to include Python.h firstAndrew Tridgell15-21/+15
If we don't include Python.h first then we get a pile of warnings due to broken redefines of XOPEN_SOURCE in the Python includes.
2009-10-23s4:dsdb 'attrs' must be static (otherwise segv with async)Andrew Bartlett1-1/+2
The async code makes this really easy to mess up... Andrew Bartlett
2009-10-23s4:dsdb Fix samba3sam test again.Andrew Bartlett1-6/+9
We again need to be careful not to call 'ldb_next_request' based functions in the partitions module. Or, we need to instead go back to having that work, and ditch the partition_request stuff... Andrew Bartlett
2009-10-23s4:dsdb Add error string in 'no such object' because of 0 replies caseAndrew Bartlett1-0/+3
2009-10-23s4:dsdb Remove unused variablesAndrew Bartlett1-2/+0
2009-10-23s4:dsdb Do less allocation when searching for partitions modulesAndrew Bartlett1-8/+10
(it didn't help that the previous allocation was on the wrong long-term context) Andrew Bartlett
2009-10-23s4:setup Mark 'cn' in secrets as case insensitiveAndrew Bartlett1-0/+1
While this does not matter very much, others may later expect 'cn' to be case insensitive. Andrew Bartlett
2009-10-23s4:secrets Look for LDAP secret with a name that is indexedAndrew Bartlett1-1/+1
This avoids a very common unindexed lookup
2009-10-23s4:gensec Use an index on computerName in schannel.ldbAndrew Bartlett1-1/+4
2009-10-23s4:dsdb Split schema loading and schema data managementAndrew Bartlett4-233/+316
By splitting the module this way, we can load the schema at startup, after the partitions module is operational, but we leave the 'mess with details of entries in the partitions' module to operate only on the partitions module. Loading the schema later allows us to set the @ATTRIBUTES correctly on all the databases. Andrew Bartlett
2009-10-23s4:dsdb Set partitions metadata as soon as it is set up.Andrew Bartlett1-3/+3
2009-10-23s4-selftest: lower some of the timeouts during make testAndrew Tridgell1-2/+2
This speeds up some of the delay based tests by a lot. There is no need to have long delays during testing.
2009-10-23s4-samdb: make sure the static credentials are never freedAndrew Tridgell1-0/+9
2009-10-23s4-ldb: move the tdb_reopen_all() calls to ldb_wrap.cAndrew Tridgell3-14/+9
2009-10-23s4-ldb: use ldb_wrap_fork_hook() to cancel child transactionsAndrew Tridgell2-2/+6
2009-10-23s4-ldb: added ldb_transaction_cancel_noerr()Andrew Tridgell3-1/+20
This will be used to allow cancelling of transactions in a child after a fork()
2009-10-23s4-server: call the ldb_wrap_fork_hook() after a fork()Andrew Tridgell2-0/+5
This will be used to allow us to cancel any pending transactions after a fork.
2009-10-23s4-server: pre-open the main ldb databases in the serverAndrew Tridgell1-8/+14
By pre-opening these databases and leaving them open, we allow the new ldb_wrap_connect() code to share the ldb context between users.
2009-10-23s4-ldbwrap: added re-use of ldb contexts in ldb_wrap_connect()Andrew Tridgell20-50/+119
This allows us to reuse a ldb context if it is open twice, instead of going through the expensive process of a full ldb open. We can reuse it if all of the parameters are the same. The change relies on callers using talloc_unlink() or free of a parent to close a ldb context.
2009-10-23s4-dsdb: add a static samdb_credentialsAndrew Tridgell2-7/+19
Similarly to system_session(), this creates a static samdb_credentials()
2009-10-23s4-dsdb: create a static system_session contextAndrew Tridgell22-39/+54
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
2009-10-22s4-lsa: fixed breakage of lsa serverAndrew Tridgell1-0/+1
2009-10-22s4-torture: fixed double free in libnet_group testAndrew Tridgell1-1/+0
2009-10-22s4:provision - Reintroduce accidentally removed parameters and checks from ↵Matthias Dieter Wallnöfer1-13/+38
"guess_names" abartlet pointed out that those are essential for a safe and consistent provision.
2009-10-22s4-lsa: fixed the lsa server to cope with the new tests from gdAndrew Tridgell1-12/+33
2009-10-22s4-ldb: added a TODO about checking the indexlistAndrew Tridgell1-0/+2
2009-10-22s4-ldb: fixed some memory leaks in new indexing codeAndrew Tridgell1-1/+3
2009-10-22s4-ldb: don't try to index non-indexed attributesAndrew Tridgell1-0/+4
2009-10-22s4-selftest: removed raw.unlink from quicktestAndrew Tridgell1-1/+0
This test takes 40s, and quicktest already covers delete operations in base.delete
2009-10-22s4-ldb: ensure new dn_list elements are not owned by callerAndrew Tridgell1-1/+1
2009-10-22s4-ldb: over-allocate index records to save on realloc costsAndrew Tridgell1-1/+5
2009-10-22s4-ldb: fixed tdb error handling in ldb_index.cAndrew Tridgell1-3/+8
2009-10-22s4-ldb: delete empty index recordsAndrew Tridgell1-0/+8
2009-10-22s4-ldb: do more validation of idxptr listsAndrew Tridgell1-12/+31
2009-10-22s4-ldb: expose ltdb_err_map and ltdb_delete_noindexAndrew Tridgell2-2/+5
These will be used by ldb_index.c
2009-10-22s4-ldb: fast path for equal pointersAndrew Tridgell1-1/+1
We compare identical ldb_val values surprisingly often
2009-10-22s4-ldb: when taking a list intersection, the result can be as long as the ↵Andrew Tridgell1-1/+1
first list Intuitively you would think it couldn't be longer than the minimum of the two lists, but we are deliberately allowing for duplicates at this level of the indexing code, which means the result can be longer
2009-10-22s4-ldb: ldb indexing rewrite - part1Andrew Tridgell4-909/+739
This gets rid of the @IDXPTR approach to in-transaction indexing, instead using an in-memory tdb to hold index values during a transaction. This also cleans up a lot of the internal indexing logic, hopefully making it easier to understand. One of the big changes is in memory management, with a lot more use made of talloc tricks to avoid copying dn lists, and shortcuts used to avoid high intersection and union calculation costs. The overall result is that a re-provision on my laptop goes from 48s to a bit over 10s.