Howard Chu
e131aa424b
ITS#4108 fix another accept/read race condition
2005-10-29 15:16:35 +00:00
Howard Chu
4bb234bac8
ITS#4108 Cleanup accept/close race conditions.
2005-10-29 04:17:56 +00:00
Howard Chu
638b84db6e
Cleanup, fix ITS#4120
2005-10-29 02:06:32 +00:00
Kurt Zeilenga
5c125d19a9
Revise last commit
2005-10-16 20:38:48 +00:00
Kurt Zeilenga
964aa48c37
Fix TLS/SASL handling in new lightweight dispatcher
2005-10-16 20:33:01 +00:00
Kurt Zeilenga
24efc30873
Move slapd_set_read before return
...
(likely wake should be after... hmmm....)
2005-10-16 17:36:30 +00:00
Kurt Zeilenga
533394d641
disable SLAP_SEM_LOAD_CONTROL by default, needs additional review/work
...
Misc. other cleanup
2005-10-16 00:17:58 +00:00
Kurt Zeilenga
9bc19c5232
Revised lightweight dispatcher -- writers moved into worker threads
...
Presently spawns operations in separate thread to ease clr_read/wake handling.
Seems to perform reasonable well...
2005-10-15 06:54:55 +00:00
Kurt Zeilenga
577bfc477e
More cleanup
2005-10-14 22:04:03 +00:00
Kurt Zeilenga
40ff060f83
cleanup
2005-10-14 21:37:14 +00:00
Howard Chu
7dcb68f6fa
Move CSN invocations to backends
2005-10-13 11:58:44 +00:00
Kurt Zeilenga
4d0de8b3c5
cleanup
2005-10-13 04:33:01 +00:00
Kurt Zeilenga
5085d1f8ff
cleanup
2005-10-13 02:08:14 +00:00
Kurt Zeilenga
6db7e605ee
Align Operation buffers using LBER_ALIGNED_BUFFER
2005-10-12 23:17:28 +00:00
Hallvard Furuseth
27831fcde5
Include "fd=.." in Statslog message "TLS established"
2005-08-29 23:34:42 +00:00
Howard Chu
3fc2e1388a
ITS#3983 fix connection_client_stop, must reset sockbuf
2005-08-29 23:05:32 +00:00
Hallvard Furuseth
9873eb7ab0
Add missing Statslog() statements (loglevel stats/stats2):
...
"ABANDON", "STARTTLS", "CANCEL", "WHOAMI", "PASSMOD", "EXT", "INTERM",
"TLS established", SASL and Extended Response "RESULT".
In Statslog output "conn=xx fd=yy closed", append the reason in
"()" unless client or server closed the connection after Unbind.
Still missing Statslog output from a number of failed requests.
2005-08-14 00:14:58 +00:00
Pierangelo Masarati
4ceeea4efa
should fix the invalid free issue
2005-07-21 19:14:09 +00:00
Luke Howard
ffe20229dc
Move most of SLAPI frontend into overlay
2005-07-21 16:35:20 +00:00
Howard Chu
143635437a
ITS#3549 don't queue Unbind requests
2005-06-18 09:06:12 +00:00
Howard Chu
0f4a5f7445
ITS#3671 must release conn->c_mutex to allow blocked writers to exit
2005-04-28 18:36:08 +00:00
Kurt Zeilenga
3e800f20bd
<limits.h> cleanup && ITS#3643
2005-04-10 19:32:14 +00:00
Howard Chu
bf0373b486
Add conn->c_sasl_authz_dn, store SASL authzDN separately from authcDN
2005-04-03 09:38:53 +00:00
Howard Chu
6cbe778c81
ITS#3549 allow one op to progress when a socket with writewaiters
...
becomes ready to write.
2005-02-18 04:21:59 +00:00
Howard Chu
5d8ece605e
ITS#3549 cleanup abandon/cancel processing
2005-02-18 01:01:35 +00:00
Howard Chu
9f30e1e650
Delete obsolete syncprovider stuff
2005-01-14 05:19:51 +00:00
Howard Chu
b6fbfe8540
ITS#3463 additional fix
2005-01-14 02:08:02 +00:00
Kurt Zeilenga
dc0eacd40b
Happy New Year!
2005-01-01 20:49:32 +00:00
Pierangelo Masarati
a8547f5945
fix comment
2005-01-01 16:14:11 +00:00
Pierangelo Masarati
4b8e57e11f
rework operations in a table-driven manner; better exception handling
2004-12-29 12:31:06 +00:00
Pierangelo Masarati
564d1a6aff
fix previous commit (for compilers that require at least one statement after a label)
2004-12-28 22:48:43 +00:00
Kurt Zeilenga
4a618959bb
PROTOS fix
2004-12-28 19:05:52 +00:00
Pierangelo Masarati
2f5e17d056
improve usability of global ovrlays for write operations; may need to anticipate SLAPI accordingly. More work needed for add (overlays need to understand they're global and call slap_mods2entry() explicitly)
2004-12-11 17:48:15 +00:00
Pierangelo Masarati
b498a9222a
better fix to ITS#3433
2004-12-11 15:41:50 +00:00
Howard Chu
948fa42499
Fix prev commit
2004-12-09 21:49:24 +00:00
Howard Chu
423c53d267
Add fe_op_abandon, call it in connection_abandon()
2004-12-09 21:17:29 +00:00
Howard Chu
5f99f7c991
More migration of syncrepl from frontend to syncprov overlay
...
define OPERATION_BUFFER_SIZE for dummy Operations
2004-11-27 00:57:37 +00:00
Howard Chu
90cc409325
Split Operation into Opheader and op
2004-11-25 22:59:00 +00:00
Howard Chu
3a5bde98ba
Disable back-bdb native syncrepl support, enable syncprov overlay,
...
remove syncrepl stuff from Operation struct
2004-11-25 21:16:54 +00:00
Jong Hyuk Choi
7fa860ef16
Add the omitted part from the original buddy allocator commit
2004-11-24 17:46:47 +00:00
Howard Chu
47dd8d8732
Dynamic control management
2004-11-23 13:08:45 +00:00
Howard Chu
7e87f54716
Revert ITS#3353 patch, it needs to be reworked.
2004-10-06 05:51:38 +00:00
Howard Chu
fda3d6260e
ITS#3353 consolidate slapd globals into a single struct
2004-10-01 07:33:16 +00:00
Pierangelo Masarati
d026e2c9f7
preformat "conn=%lu op=%lu"
2004-09-26 23:00:00 +00:00
Pierangelo Masarati
761f287943
multiple precision with BIGNUM/gmp/ulong
2004-09-26 22:58:47 +00:00
Pierangelo Masarati
55e251e6e8
lots of cleanup; few improvements; fix RDN selection bug when creating connection dynamic entries; rework entries ordering - now it's more intuitive and regular; optimize counters update
2004-09-12 22:12:58 +00:00
Pierangelo Masarati
a637926bf2
use GMP for multiple precision in counters
2004-09-11 16:15:53 +00:00
Kurt Zeilenga
d611a4b49a
unifdef -UNEW_LOGGING
2004-09-04 04:54:28 +00:00
Luke Howard
2316461b93
Remove unused variable
2004-08-25 12:31:43 +00:00
Luke Howard
ea6f5bad65
This patch fixes some subtle interactions between SLAPI and syncrepl. Due to
...
SLAPI always assigning connection and operation IDs of zero for internal
operations, such operations would cause a stale contextCSN to be returned from
slap_get_commit_csn(). As a result, SLAPI internal updates would be invisible
to replicas until an external update was made. Also, SLAPI internal operations
never called slap_graduate_commit_csn() which leaked pending CSNs.
Also included in this patch is a general cleanup of some of the SLAPI code.
Note that we need to use a separate mutex on conn_nextid to avoid a deadlock
where a post-operation plugin tries to acquire connections_mutex, having locked
the per-connection mutex, while the listener thread tries to acquire the
per-connection mutex (having locked connections_mutex). connection.c needs to
be fixed to acquire mutexes in the same order.
2004-08-25 11:52:55 +00:00
