Commit Graph

8846 Commits

Author SHA1 Message Date
Howard Chu
354e90a0d9 From rev 1.373, releasing the mutex means connection_close() may get called
from another thread. Drop the assert, just return instead.
2006-10-17 19:31:42 +00:00
Pierangelo Masarati
40a893a07e temporarily change umask only if required 2006-10-15 18:28:22 +00:00
Pierangelo Masarati
8567ce8eed fix previous commit 2006-10-15 18:24:30 +00:00
Pierangelo Masarati
b6525c14d9 missing decl 2006-10-15 18:03:25 +00:00
Pierangelo Masarati
c0be820648 fix format; silence warning 2006-10-15 18:01:27 +00:00
Pierangelo Masarati
cbf05fd777 log messages related to blocking failures at "none" level as well, so they make it through slaptest 2006-10-15 15:03:02 +00:00
Pierangelo Masarati
59e60483fa create IPC sockets with wide-open permissions (0222 might suffice; ITS#4709) 2006-10-14 17:35:50 +00:00
Pierangelo Masarati
f449ee6500 cleanup and streamline response handling 2006-10-14 17:12:05 +00:00
Pierangelo Masarati
968da1cb14 add LDAP_OPT_DIAGNOSTIC_MESSAGE; replace deprecated option names 2006-10-14 17:04:17 +00:00
Pierangelo Masarati
b1a2be99c6 don't mix return codes (ITS#4711) 2006-10-14 09:26:22 +00:00
Pierangelo Masarati
fb2d3926ce only abort update if failed pre/post read control is critical: the only thing RFC 4527 says about failure is that in case of failure of the read, no control is appended to the result message; maybe no failure should ever occur (please review) 2006-10-14 01:08:41 +00:00
Pierangelo Masarati
117caa9aeb read entry should allow special attributes in request (*, +, 1.1) 2006-10-14 00:59:55 +00:00
Pierangelo Masarati
07f61475c7 read entry needs o_bd for ACLs 2006-10-14 00:59:22 +00:00
Howard Chu
6f13331f4f Drop extraneous test from copy/paste (1.200) (coverity) 2006-10-08 00:37:18 +00:00
Howard Chu
65f7c1a50d Fix uninit (coverity) 2006-10-08 00:31:08 +00:00
Howard Chu
acdddee3ab Fix uninit'd cb (coverity) 2006-10-08 00:26:32 +00:00
Pierangelo Masarati
f8d1a61866 more on clarification of special proxy identities and resilience to connection failure 2006-10-07 18:07:56 +00:00
Pierangelo Masarati
eb29c7c37b reword comment 2006-10-07 13:12:16 +00:00
Howard Chu
b14e6e00b5 ITS#4695 same fix for delta-sync 2006-10-06 01:57:50 +00:00
Howard Chu
4e9c39d626 ITS#4695 fix modrdn newSup 2006-10-05 03:56:23 +00:00
Kurt Zeilenga
d39809f1ab Incorporate a bit of text from RFC 4524, just to make a point regarding
ITS#4693.
2006-10-03 21:23:02 +00:00
Howard Chu
e9ecaa5d81 ITS#4692 entries without pwdChangedTime attribute have non-expiring pw 2006-10-03 09:25:54 +00:00
Pierangelo Masarati
5714f8565f fix rebind with idassert when a connection expires; also, fix override 2006-09-30 01:11:19 +00:00
Pierangelo Masarati
576646abb1 DB_BUFFER_SMALL is BDB 4.4 only 2006-09-27 12:34:00 +00:00
Ralf Haferkamp
e1f068f712 Check for the correct return code of cursor->c_get() (DB_BUFFER_SMALL) 2006-09-27 10:51:06 +00:00
Pierangelo Masarati
b769f44619 fix ITS#4686 (retry with idassert) 2006-09-26 15:12:07 +00:00
Pierangelo Masarati
0e9983ff2b fix ITS#4686 (retry with idassert) 2006-09-26 14:51:47 +00:00
Pierangelo Masarati
067d6904c2 don't know exactly the reason, but here without this back-hdb suddenly starts crashings all times (since bdb_monitor_* stuff) 2006-09-26 08:43:01 +00:00
Pierangelo Masarati
8f90eb97ec use right, although empty, function in bdb_db_destroy() 2006-09-25 23:45:29 +00:00
Pierangelo Masarati
3133b0a704 not sure, but I believe back-hdb needs monitor.c as well... 2006-09-25 23:44:50 +00:00
Pierangelo Masarati
c4a0d94883 monitor max file descriptor number 2006-09-25 23:44:17 +00:00
Howard Chu
393e6e5ec0 Tweak prev commit, avoid referencing syncop when resetting qtask 2006-09-21 10:12:42 +00:00
Howard Chu
c1fad2d603 ITS#4658 propagate errors from syncprov_sendresp; drop qtask on errors.
Note: need a way to close the connection too.
2006-09-21 09:42:40 +00:00
Howard Chu
d6d31af4f0 send_search_reference should propagate errors 2006-09-21 09:28:49 +00:00
Howard Chu
db082f4b1d Fix prev commit 2006-09-20 20:22:59 +00:00
Howard Chu
f63f46f0b1 Use a single malloc for entry_decode instead of two - partially decode
the entry to get nvals first, so the size can be computed.
2006-09-20 02:43:47 +00:00
Pierangelo Masarati
cbc238721a typo in previous commit... 2006-09-17 22:24:41 +00:00
Pierangelo Masarati
036289f9c4 support for Solaris' /dev/poll (need to manually define SLAP_X_DEVPOLL; needs work yet) 2006-09-17 22:17:34 +00:00
Pierangelo Masarati
dd4d074068 reorganize select interface (ready for /dev/poll) 2006-09-17 22:09:50 +00:00
Pierangelo Masarati
c0c089f4a4 make sure we don't interfere with frontend's entry when the suffix is empty (test027) 2006-09-17 16:45:24 +00:00
Pierangelo Masarati
7fd11ad679 minor cleanup 2006-09-17 16:44:40 +00:00
Kurt Zeilenga
7c68de27c8 Check opattrs add result 2006-09-16 17:20:29 +00:00
Kurt Zeilenga
2c4b2b5566 return structuralObjectClass errors
(likely need to be applied to other backends)
2006-09-16 11:48:52 +00:00
Pierangelo Masarati
2b242dee14 private database inherits monitoring from underlying one 2006-09-15 00:23:14 +00:00
Pierangelo Masarati
c92571680d also work with pcache 2006-09-15 00:16:55 +00:00
Pierangelo Masarati
f34d418ced more good looking error messages 2006-09-15 00:16:35 +00:00
Pierangelo Masarati
4add034d5d destroy exactly what needs to 2006-09-14 23:53:23 +00:00
Pierangelo Masarati
6fd9090a0d fix one-time leak 2006-09-14 22:50:07 +00:00
Howard Chu
f1457a8ce8 Additional fix for ITS#4673 2006-09-14 22:38:45 +00:00
Pierangelo Masarati
524a7bc27a by now, don't bail out if monitoring cannot be set up 2006-09-14 22:26:37 +00:00