]>
git.sur5r.net Git - openldap/log
Howard Chu [Wed, 13 Sep 2006 13:26:56 +0000 (13:26 +0000)]
Use register_at / register_oc
Howard Chu [Wed, 13 Sep 2006 13:05:19 +0000 (13:05 +0000)]
Use register_at / register_oc helpers
Howard Chu [Wed, 13 Sep 2006 12:48:45 +0000 (12:48 +0000)]
Add register_at / register_oc helpers.
Howard Chu [Wed, 13 Sep 2006 11:38:51 +0000 (11:38 +0000)]
Add BackendInfo.bi_extra for backend custom APIs
Use bi_extra for back-monitor entry points. Backends should never
directly reference each other's symbols.
Howard Chu [Wed, 13 Sep 2006 05:34:38 +0000 (05:34 +0000)]
Fix ITS#4672, clear errno before calling STRTOL
Howard Chu [Wed, 13 Sep 2006 05:32:42 +0000 (05:32 +0000)]
Fix typo from 1.667
Pierangelo Masarati [Tue, 12 Sep 2006 21:02:30 +0000 (21:02 +0000)]
document the syntax of the ldap.conf(5) file
Pierangelo Masarati [Tue, 12 Sep 2006 20:41:37 +0000 (20:41 +0000)]
use a more reasonable max line length (ITS#4669)
Pierangelo Masarati [Tue, 12 Sep 2006 07:53:19 +0000 (07:53 +0000)]
leave monitor calls in place, hide contents; in the long run, enable monitor and build as static by default
Howard Chu [Tue, 12 Sep 2006 02:59:28 +0000 (02:59 +0000)]
Use #ifdef SLAPD_MONITOR for monitor APIs
Howard Chu [Tue, 12 Sep 2006 00:16:52 +0000 (00:16 +0000)]
Fix xref broken by sasl.sdf 1.30
Pierangelo Masarati [Mon, 11 Sep 2006 22:45:19 +0000 (22:45 +0000)]
ldap_back_dobind_int() may need to free a connection
Howard Chu [Mon, 11 Sep 2006 21:19:38 +0000 (21:19 +0000)]
In connection_next, sockets may not be contiguous on Unix if syslog etc are
active. Just search to the end of the table, tracking the max isn't worth
the lock overhead.
Pierangelo Masarati [Mon, 11 Sep 2006 21:04:25 +0000 (21:04 +0000)]
rearrange error handling
Pierangelo Masarati [Mon, 11 Sep 2006 00:52:43 +0000 (00:52 +0000)]
add sb_version to slap_bindconf;
rework slap_client_connect() and slap_discover_feature() accordingly
Pierangelo Masarati [Mon, 11 Sep 2006 00:06:03 +0000 (00:06 +0000)]
make sure we got the right control...
Howard Chu [Sat, 9 Sep 2006 22:15:45 +0000 (22:15 +0000)]
Fix uninit'd *cb (Coverity scan)
Pierangelo Masarati [Sat, 9 Sep 2006 17:01:28 +0000 (17:01 +0000)]
ld may legitimately be NULL if another operation freed it because of a failure while binding
Pierangelo Masarati [Sat, 9 Sep 2006 16:25:32 +0000 (16:25 +0000)]
don't bother about unregistering stuff during shutdown
Pierangelo Masarati [Sat, 9 Sep 2006 16:24:35 +0000 (16:24 +0000)]
further improve monitoring setup and cleanup
Pierangelo Masarati [Sat, 9 Sep 2006 16:23:02 +0000 (16:23 +0000)]
- use STRLENOF, looks cleaner
- in case a value is incorrectly empty (technically, it's null because
that's how it's returned by the stringification function), make enough
room to add the error string "(null)"
- in case a value is legitimately empty (see above), use it accordingly;
for example, a filter "(entryDN=)" is valid, so it should be possible
to stringify it like that.
Pierangelo Masarati [Sat, 9 Sep 2006 16:20:11 +0000 (16:20 +0000)]
minor cleanup
Pierangelo Masarati [Sat, 9 Sep 2006 16:16:27 +0000 (16:16 +0000)]
rename functions/macros (no impact, they're currently unused)
Pierangelo Masarati [Sat, 9 Sep 2006 02:35:58 +0000 (02:35 +0000)]
more cleanup; add the frontend to the set of monitored databases; handle exceptional cases a bit more smoothly
Pierangelo Masarati [Sat, 9 Sep 2006 00:10:43 +0000 (00:10 +0000)]
fix entry unlock in case of unhandled scope (ITS#4668)
Pierangelo Masarati [Sat, 9 Sep 2006 00:00:44 +0000 (00:00 +0000)]
support "subordinate" scope
Howard Chu [Fri, 8 Sep 2006 23:31:53 +0000 (23:31 +0000)]
ITS#4667 in connection_closing, release c_mutex to allow waiters
to clear out.
Howard Chu [Fri, 8 Sep 2006 23:30:31 +0000 (23:30 +0000)]
ITS#4667 fix assert in connection_next() for PENDING connections
Pierangelo Masarati [Fri, 8 Sep 2006 22:10:49 +0000 (22:10 +0000)]
cleanup names
Pierangelo Masarati [Fri, 8 Sep 2006 22:02:32 +0000 (22:02 +0000)]
- need to be able to set new subsystems when appending entire entries;
- in case we appended an entry to a yet unknown parent, we don't know
the full DN, so unregistering needs to be possible based on the
available data;
- complete unregister calls so that not-yet-registered stuff can be
removed from limbo
Pierangelo Masarati [Fri, 8 Sep 2006 21:19:47 +0000 (21:19 +0000)]
some overlay (rwm?) might set those values
Pierangelo Masarati [Fri, 8 Sep 2006 21:19:05 +0000 (21:19 +0000)]
really really minor cleanup
Pierangelo Masarati [Fri, 8 Sep 2006 20:49:00 +0000 (20:49 +0000)]
defer handling of loglevel listing, to let modules register custom subsystems (ITS#4666). Note: in case of any error in between, logging could not occur
Pierangelo Masarati [Fri, 8 Sep 2006 18:26:57 +0000 (18:26 +0000)]
point to slapd(8) for details on "-d"
Pierangelo Masarati [Fri, 8 Sep 2006 18:26:20 +0000 (18:26 +0000)]
re-enable "-d '?'" as a means to list what loglevels are available (ITS#4666)
Pierangelo Masarati [Fri, 8 Sep 2006 18:25:34 +0000 (18:25 +0000)]
complain if "-T" is not the first option
Pierangelo Masarati [Fri, 8 Sep 2006 18:10:53 +0000 (18:10 +0000)]
re-enable "-d '?'" as a means to list what loglevels are available (ITS#4666)
Pierangelo Masarati [Wed, 6 Sep 2006 23:08:57 +0000 (23:08 +0000)]
custom monitoring can be enabled/disabled via config
Pierangelo Masarati [Wed, 6 Sep 2006 16:06:49 +0000 (16:06 +0000)]
cleanup subsystems monitoring OID handling
Pierangelo Masarati [Wed, 6 Sep 2006 15:12:36 +0000 (15:12 +0000)]
check return values (failing when filter invalid)
Pierangelo Masarati [Wed, 6 Sep 2006 15:11:54 +0000 (15:11 +0000)]
fix use with back-hdb
Pierangelo Masarati [Wed, 6 Sep 2006 12:07:03 +0000 (12:07 +0000)]
cleanup monitoring resources
Pierangelo Masarati [Wed, 6 Sep 2006 12:06:42 +0000 (12:06 +0000)]
implement support for unregistering registered stuff
Pierangelo Masarati [Wed, 6 Sep 2006 12:05:38 +0000 (12:05 +0000)]
improve previous commit
Pierangelo Masarati [Wed, 6 Sep 2006 12:04:52 +0000 (12:04 +0000)]
fix outdated comments
Pierangelo Masarati [Wed, 6 Sep 2006 12:04:33 +0000 (12:04 +0000)]
add scope helpers
Pierangelo Masarati [Wed, 6 Sep 2006 09:10:46 +0000 (09:10 +0000)]
- objectClass is now AUXILIARY (modifies existing object...)
- moved directory path publication from back-monitor/database.c
- added cleanup handler
- minor cleanup
Pierangelo Masarati [Wed, 6 Sep 2006 09:09:03 +0000 (09:09 +0000)]
remove dependencies on back-{b,h}db
Pierangelo Masarati [Wed, 6 Sep 2006 09:06:50 +0000 (09:06 +0000)]
add attr_normalize*() helper (please holler if I missed any obvious alternative)
Howard Chu [Wed, 6 Sep 2006 03:37:13 +0000 (03:37 +0000)]
Return current entry/IDL cache counts
Pierangelo Masarati [Wed, 6 Sep 2006 02:14:32 +0000 (02:14 +0000)]
don't get mad if monitor database is not configured
Pierangelo Masarati [Wed, 6 Sep 2006 02:06:45 +0000 (02:06 +0000)]
minor cleanup; plug couple of one-time leaks
Pierangelo Masarati [Wed, 6 Sep 2006 01:54:54 +0000 (01:54 +0000)]
- setup framework for monitoring of back-bdb/back-hdb stuff in their
database entry in cn=monitor
- improve support for custom attribute/callback registration
in back-monitor
- design API for registered antry/attr/callback removal in back-monitor
HEADS UP: now back-bdb requires back-monitor to be present, either static
or dynamically loaded __before__ back-bdb/back-hdb is initialized.
This works for sure if both back-bdb & back-monitor are static or dynamic
(provided back-monitor.la is loaded before back_bdb.la); it is not clear
what happens if back-bdb is static and back-monitor is dynamic. In case,
either make back-bdb monitoring conditional, or hardwire back-monitor
much like back-config and back-ldif.
Pierangelo Masarati [Mon, 4 Sep 2006 08:40:40 +0000 (08:40 +0000)]
better handling of randomness
Pierangelo Masarati [Mon, 4 Sep 2006 08:36:16 +0000 (08:36 +0000)]
cleanup handling of sleeps
Pierangelo Masarati [Mon, 4 Sep 2006 08:29:47 +0000 (08:29 +0000)]
append logs
Pierangelo Masarati [Mon, 4 Sep 2006 08:26:09 +0000 (08:26 +0000)]
improve timeout support (ITS#4157, ITS#4663)
Pierangelo Masarati [Mon, 4 Sep 2006 08:24:05 +0000 (08:24 +0000)]
improve timeout support (ITS#4157, ITS#4663); manpage cleanup
Pierangelo Masarati [Mon, 4 Sep 2006 07:20:59 +0000 (07:20 +0000)]
release authz resources at shutdown
Pierangelo Masarati [Mon, 4 Sep 2006 07:17:34 +0000 (07:17 +0000)]
move req2op selection into a helper
Pierangelo Masarati [Sun, 3 Sep 2006 11:14:37 +0000 (11:14 +0000)]
fix arg format; better trace logging...
Pierangelo Masarati [Sun, 3 Sep 2006 11:04:27 +0000 (11:04 +0000)]
don't abandon binds in progress; rather unbind (ITS#4663).
Better logging for tracing
Pierangelo Masarati [Sun, 3 Sep 2006 11:02:15 +0000 (11:02 +0000)]
back-ldap should only touch the connection timestamp when activity actually occurred (i.e. ldap_result() returned a positive value), otherwise if "the network cable is removed after the connection is established" ldap_result() keeps returning 0 forever...
Pierangelo Masarati [Sun, 3 Sep 2006 10:58:13 +0000 (10:58 +0000)]
use preallocated entries? (otherwise bdb_add and others leak...)
Pierangelo Masarati [Sun, 3 Sep 2006 10:53:38 +0000 (10:53 +0000)]
plug slapo-ppolicy related leaks (ITS#4665)
Pierangelo Masarati [Sun, 3 Sep 2006 10:45:07 +0000 (10:45 +0000)]
plug leaks in search filter (and cleanup code; ITS#4664)
Pierangelo Masarati [Fri, 1 Sep 2006 16:21:20 +0000 (16:21 +0000)]
better function name
Pierangelo Masarati [Fri, 1 Sep 2006 14:00:58 +0000 (14:00 +0000)]
keep count of requests in use
Pierangelo Masarati [Fri, 1 Sep 2006 13:59:55 +0000 (13:59 +0000)]
don't risk dangling pointers
Pierangelo Masarati [Fri, 1 Sep 2006 13:57:37 +0000 (13:57 +0000)]
don't use deprecated functions
Pierangelo Masarati [Fri, 1 Sep 2006 10:41:54 +0000 (10:41 +0000)]
remove unnecessary assertion; add test for pending ops
Pierangelo Masarati [Fri, 1 Sep 2006 10:01:36 +0000 (10:01 +0000)]
cleanup
Pierangelo Masarati [Fri, 1 Sep 2006 10:00:50 +0000 (10:00 +0000)]
add an assertion to track ITS#4659; minor cleanup
Pierangelo Masarati [Thu, 31 Aug 2006 17:34:01 +0000 (17:34 +0000)]
minor cleanup
Pierangelo Masarati [Thu, 31 Aug 2006 16:21:54 +0000 (16:21 +0000)]
don't leak identity
Pierangelo Masarati [Thu, 31 Aug 2006 13:51:25 +0000 (13:51 +0000)]
destroy socket set (man page for epoll_create(2) states that close(2) must be called on the returned filedes)
Pierangelo Masarati [Thu, 31 Aug 2006 13:35:01 +0000 (13:35 +0000)]
fix typo
Pierangelo Masarati [Wed, 30 Aug 2006 22:45:26 +0000 (22:45 +0000)]
pick declaration of ch_strdup()
Pierangelo Masarati [Wed, 30 Aug 2006 21:05:05 +0000 (21:05 +0000)]
don't run useless (and likely to fail) checks, if possible
Pierangelo Masarati [Tue, 29 Aug 2006 17:09:08 +0000 (17:09 +0000)]
fix pre-allocated entry/attribute handling
Pierangelo Masarati [Tue, 29 Aug 2006 16:07:39 +0000 (16:07 +0000)]
fix pre-allocated entry/attribute handling
Kurt Zeilenga [Tue, 29 Aug 2006 07:38:09 +0000 (07:38 +0000)]
forced commit
Kurt Zeilenga [Tue, 29 Aug 2006 07:28:22 +0000 (07:28 +0000)]
more cleanup
Kurt Zeilenga [Tue, 29 Aug 2006 07:25:34 +0000 (07:25 +0000)]
cleanup
Kurt Zeilenga [Tue, 29 Aug 2006 07:22:28 +0000 (07:22 +0000)]
minor change
Howard Chu [Tue, 29 Aug 2006 02:02:12 +0000 (02:02 +0000)]
Hide config schema if !LDAP_DEVEL
Howard Chu [Tue, 29 Aug 2006 01:43:23 +0000 (01:43 +0000)]
Entry/Attribute struct caching, to minimize malloc fragmentation
Note: this breaks LDAP_COMP_MATCH and SLAP_ZONE_MALLOC. But they
were probably broken already anyway.
Howard Chu [Tue, 29 Aug 2006 01:36:52 +0000 (01:36 +0000)]
Must use ch_strdup, not strdup
Howard Chu [Tue, 29 Aug 2006 01:34:55 +0000 (01:34 +0000)]
TLS ctx is already cleaned up
Howard Chu [Tue, 29 Aug 2006 01:19:34 +0000 (01:19 +0000)]
Only set c_sasl_bindop on actual SASL binds
Pierangelo Masarati [Mon, 28 Aug 2006 14:33:59 +0000 (14:33 +0000)]
note the single-value userPassword constraint (please review)
Pierangelo Masarati [Mon, 28 Aug 2006 13:31:16 +0000 (13:31 +0000)]
append global access rules also when be != NULL in backend_startup() (ITS#4654)
Pierangelo Masarati [Mon, 28 Aug 2006 12:44:39 +0000 (12:44 +0000)]
Add single-password constraint; NOTE from ppolicy.c:
/* FIXME: there's no easy way to ensure
* that add does not cause multiple
* userPassword values; one way (that
* would be consistent with the single
* password constraint) would be to turn
* add into replace); another would be
* to disallow add.
*
* Let's check at least that a single value
* is being added
*/
Quanah Gibson-Mount [Mon, 28 Aug 2006 01:07:25 +0000 (01:07 +0000)]
ITS#4358, thanks to Kevine Zerbib
Quanah Gibson-Mount [Mon, 28 Aug 2006 01:05:32 +0000 (01:05 +0000)]
Fix comment style.
Quanah Gibson-Mount [Mon, 28 Aug 2006 01:01:32 +0000 (01:01 +0000)]
Fix delete type modifys where no value is given to be deleted. (ITS#2612)
Thanks to Alister Winfield for the patch.
Quanah Gibson-Mount [Sat, 26 Aug 2006 23:20:00 +0000 (23:20 +0000)]
fix e_private value, goes along with memory leak fix to accesslog.c
Pierangelo Masarati [Sat, 26 Aug 2006 17:55:56 +0000 (17:55 +0000)]
reduce code duplication
Pierangelo Masarati [Sat, 26 Aug 2006 17:28:07 +0000 (17:28 +0000)]
disable X-ORDERED 'VALUES' naming attributes by now
Pierangelo Masarati [Sat, 26 Aug 2006 15:24:49 +0000 (15:24 +0000)]
invalidate idle connection if a candidate target does not respond for the duration of time limit...