]>
git.sur5r.net Git - openldap/log
Howard Chu [Fri, 7 Oct 2011 02:15:10 +0000 (19:15 -0700)]
More for cursor reuse
Howard Chu [Fri, 7 Oct 2011 02:14:22 +0000 (19:14 -0700)]
Keep some dn2i/id2e cursors open longer
So they can be reused, and avoid unnecessary page_searches
Howard Chu [Tue, 4 Oct 2011 09:18:09 +0000 (02:18 -0700)]
Turn off tool IDL caching
Well, just like in back-bdb in 2005, it only slowed things down
on larger LDIFs.
Howard Chu [Mon, 3 Oct 2011 18:29:04 +0000 (11:29 -0700)]
Tweak tool IDL cache, use MDB_MULTIPLE
Howard Chu [Mon, 3 Oct 2011 04:29:21 +0000 (21:29 -0700)]
Use MDB_APPEND mode where possible
Howard Chu [Mon, 3 Oct 2011 02:31:18 +0000 (19:31 -0700)]
Use IDL caching in tool/quick mode
Howard Chu [Sun, 2 Oct 2011 17:06:06 +0000 (10:06 -0700)]
Use MDB_RESERVE for id2entry_put
Howard Chu [Wed, 21 Sep 2011 22:59:08 +0000 (15:59 -0700)]
Updated for mdb_txn_begin API change.
Nested txn stuff which we don't use.
Hallvard Furuseth [Tue, 13 Sep 2011 07:39:28 +0000 (09:39 +0200)]
back-bdb/back-mdb IDL cleanup.
Parenthesize macros. Microoptimize IDL search. Use RANGE_<FIRST/LAST>
when IDL is known to be a range.
Hallvard Furuseth [Sun, 18 Sep 2011 19:49:25 +0000 (21:49 +0200)]
Cleanup ISO C compatibility for recent commits
Hallvard Furuseth [Sun, 18 Sep 2011 20:37:45 +0000 (22:37 +0200)]
More libmdb vs back-mdb C type tweaks.
Do not access a size_t via an unsigned long pointer, an MDB_val via a
berval pointer, or the inverse. Instead copy into the proper data types.
Howard Chu [Sun, 18 Sep 2011 21:47:41 +0000 (14:47 -0700)]
Fix prev commit
Don't truncate all indices if specific indices were given.
Howard Chu [Sun, 18 Sep 2011 21:28:36 +0000 (14:28 -0700)]
Support index truncation
Howard Chu [Sun, 18 Sep 2011 20:25:38 +0000 (13:25 -0700)]
New mdb_close() signature
Howard Chu [Wed, 14 Sep 2011 17:14:02 +0000 (10:14 -0700)]
Fix entry_alloc with no attrs
Howard Chu [Wed, 14 Sep 2011 16:54:58 +0000 (09:54 -0700)]
Fix onelevel search from last commit
Howard Chu [Tue, 13 Sep 2011 23:58:38 +0000 (16:58 -0700)]
More search optimization
Tighten up entry_alloc/entry_decode
Track parent nodes in idscopes
Howard Chu [Tue, 13 Sep 2011 22:19:04 +0000 (15:19 -0700)]
Tentative speedup - use plain ints for entry_encode/decode
Keep entries word-aligned. Yields ~10% speedup, but also 10% more space used.
May revert this if the space cost is unacceptable.
Howard Chu [Tue, 13 Sep 2011 08:28:37 +0000 (01:28 -0700)]
mdb_db_close() should not force env_sync in quick mode
Quanah Gibson-Mount [Fri, 4 Nov 2011 21:44:01 +0000 (14:44 -0700)]
Resolve MDB/HDB monitor OID clashes
Conflicts:
servers/slapd/bconfig.c
Howard Chu [Mon, 12 Sep 2011 18:05:38 +0000 (11:05 -0700)]
Just use memcpy since we know it's non-overlapping
Hallvard Furuseth [Mon, 12 Sep 2011 13:58:43 +0000 (15:58 +0200)]
Tweak back-mdb integer types to match libmdb.
Also do not cast between berval and MDB_val. That breaks when
sizeof(berval.bv_len) != sizeof(MDB_val.mv_size).
Howard Chu [Sat, 10 Sep 2011 08:12:50 +0000 (01:12 -0700)]
Fix - slapcat doesn't set op->o_hdr for tmpmemctx
Howard Chu [Sat, 10 Sep 2011 05:00:34 +0000 (22:00 -0700)]
Fix out-of-order slapadd
Howard Chu [Sat, 10 Sep 2011 02:53:56 +0000 (19:53 -0700)]
Fix bogus error codes
Howard Chu [Fri, 9 Sep 2011 07:10:48 +0000 (00:10 -0700)]
Keep cursor open for id2entry
Howard Chu [Fri, 9 Sep 2011 05:07:04 +0000 (22:07 -0700)]
Use tmpmem-based entries
Howard Chu [Fri, 9 Sep 2011 05:02:59 +0000 (22:02 -0700)]
Drop referral.c
Howard Chu [Thu, 8 Sep 2011 17:31:04 +0000 (10:31 -0700)]
Howard Chu [Wed, 7 Sep 2011 00:23:41 +0000 (17:23 -0700)]
More error msgs
Howard Chu [Tue, 6 Sep 2011 21:04:38 +0000 (14:04 -0700)]
Fix
6c8e4f2 for empty suffixes
Howard Chu [Tue, 6 Sep 2011 09:21:19 +0000 (02:21 -0700)]
More search optimizations
Howard Chu [Tue, 6 Sep 2011 02:49:35 +0000 (19:49 -0700)]
Use AD-indexes
Howard Chu [Mon, 5 Sep 2011 22:23:08 +0000 (15:23 -0700)]
Debug helper for index
Hallvard Furuseth [Mon, 5 Sep 2011 19:10:04 +0000 (21:10 +0200)]
back-mdb/Makefile.in += variable MDB_SUBDIR.
For compiling against a checkout of the stand-alone libmdb,
which has a branch 'mdb.master' with its own root:
git clone --branch=mdb.master <openldap_repo_dir> mdb
set_mdb="MDB_SUBDIR=`pwd`/mdb/libraries/libmdb"
... maybe modify mdb repository for testing ...
cd <openldap_repo_dir>; make $set_mdb
Howard Chu [Mon, 5 Sep 2011 10:52:38 +0000 (03:52 -0700)]
Fix from key_change consolidation
Howard Chu [Mon, 5 Sep 2011 08:23:35 +0000 (01:23 -0700)]
Cursor updates
Howard Chu [Sat, 3 Sep 2011 20:09:19 +0000 (13:09 -0700)]
Fix conversion of IDL to range
Howard Chu [Thu, 1 Sep 2011 23:52:53 +0000 (16:52 -0700)]
Import back-mdb
Quanah Gibson-Mount [Fri, 4 Nov 2011 21:06:23 +0000 (14:06 -0700)]
Merge remote branch 'origin/mdb.master' into OPENLDAP_REL_ENG_2_4
Quanah Gibson-Mount [Fri, 4 Nov 2011 20:27:38 +0000 (13:27 -0700)]
ITS#6848
Howard Chu [Tue, 1 Mar 2011 01:31:11 +0000 (01:31 +0000)]
More for #6848 - drop -w option, just make it default behavior
Howard Chu [Tue, 1 Mar 2011 01:19:37 +0000 (01:19 +0000)]
ITS#6848 Add -w option to wait for DB startup before parent exits
Quanah Gibson-Mount [Fri, 4 Nov 2011 20:22:54 +0000 (13:22 -0700)]
ITS#6666
Howard Chu [Fri, 4 Feb 2011 17:36:49 +0000 (17:36 +0000)]
Overlay result is in rs->sr_err
Howard Chu [Fri, 4 Feb 2011 10:52:34 +0000 (10:52 +0000)]
Add overlay config stuff, handle overlay responses
Howard Chu [Thu, 3 Feb 2011 21:54:11 +0000 (21:54 +0000)]
Just use back-sock OID arc for overlay
Howard Chu [Thu, 3 Feb 2011 21:34:45 +0000 (21:34 +0000)]
As an overlay, always return SLAP_CB_CONTINUE
Howard Chu [Thu, 3 Feb 2011 21:24:26 +0000 (21:24 +0000)]
Add response message for overlays
Howard Chu [Thu, 3 Feb 2011 20:47:38 +0000 (20:47 +0000)]
More for overlay use, add connid extension
Howard Chu [Thu, 3 Feb 2011 20:17:23 +0000 (20:17 +0000)]
Add overlay wrapper, use for triggers etc.
Quanah Gibson-Mount [Fri, 4 Nov 2011 19:30:11 +0000 (12:30 -0700)]
ITS#6993
Pierangelo Masarati [Mon, 22 Aug 2011 17:23:01 +0000 (11:23 -0600)]
release mutex only *after* backend connection initialization (ITS#6993)
Quanah Gibson-Mount [Fri, 4 Nov 2011 19:05:17 +0000 (12:05 -0700)]
ITS#6826
Howard Chu [Mon, 7 Feb 2011 01:14:04 +0000 (01:14 +0000)]
Add CVS ID
Howard Chu [Mon, 7 Feb 2011 01:12:49 +0000 (01:12 +0000)]
Drop unneeded scheme param
Howard Chu [Mon, 7 Feb 2011 01:09:47 +0000 (01:09 +0000)]
Add APR1 to Makefile/README. Add {BSDMD5} mechanism.
Howard Chu [Mon, 7 Feb 2011 00:47:25 +0000 (00:47 +0000)]
Replace pw_string64 with lutil_passwd_string64
Howard Chu [Mon, 7 Feb 2011 00:46:06 +0000 (00:46 +0000)]
Export pw_string64 as lutil_passwd_string64
Howard Chu [Mon, 7 Feb 2011 00:49:55 +0000 (00:49 +0000)]
Use lutil_passwd_string64 instead of pw_string64
Howard Chu [Mon, 7 Feb 2011 00:49:04 +0000 (00:49 +0000)]
ITS#6826 from Devin J. Pohly
Quanah Gibson-Mount [Fri, 4 Nov 2011 18:58:46 +0000 (11:58 -0700)]
ITS#3272
Howard Chu [Thu, 17 Feb 2011 02:57:47 +0000 (02:57 +0000)]
ITS#3272 link against libslapi.la
Howard Chu [Thu, 17 Feb 2011 02:56:06 +0000 (02:56 +0000)]
LIBSLAPITOOLS is obsolete since tools subdir is long gone
ITS#3272 link to libslapi.la
Howard Chu [Fri, 4 Nov 2011 18:32:38 +0000 (11:32 -0700)]
ITS#7081 fix prev commit
Quanah Gibson-Mount [Thu, 3 Nov 2011 23:13:48 +0000 (16:13 -0700)]
ITS#6981
Howard Chu [Wed, 2 Nov 2011 16:26:50 +0000 (09:26 -0700)]
ITS#6981 pcache: don't reference rs->sr_entry in op_cleanup
Howard Chu [Wed, 2 Nov 2011 16:25:45 +0000 (09:25 -0700)]
Revert "ITS#6981 don't free entries prematurely"
This reverts commit
877ebdf5ef7c40e9d8aeba3894cd07f95bc721d2 .
Howard Chu [Tue, 1 Nov 2011 23:24:10 +0000 (16:24 -0700)]
ITS#6981 don't free entries prematurely
Howard Chu [Tue, 1 Nov 2011 23:14:12 +0000 (16:14 -0700)]
Revert "ITS#6981 don't flush entries until after cleanup callbacks"
This reverts commit
b68fa5ecd73037ebd436a2663003f544d482f71e .
Howard Chu [Tue, 1 Nov 2011 20:17:06 +0000 (13:17 -0700)]
ITS#6981 don't flush entries until after cleanup callbacks
Quanah Gibson-Mount [Thu, 3 Nov 2011 23:12:14 +0000 (16:12 -0700)]
ITS#7060
Howard Chu [Tue, 1 Nov 2011 19:59:36 +0000 (12:59 -0700)]
ITS#7060 fix str2result return code parsing
Quanah Gibson-Mount [Thu, 3 Nov 2011 23:11:06 +0000 (16:11 -0700)]
ITS#7062
Howard Chu [Tue, 1 Nov 2011 18:02:00 +0000 (11:02 -0700)]
ITS#7062 call response callbacks even if connection was dropped
Howard Chu [Thu, 13 Oct 2011 02:19:52 +0000 (19:19 -0700)]
Define S_IWRITE if needed
Howard Chu [Thu, 13 Oct 2011 02:18:05 +0000 (19:18 -0700)]
Android defines wctomb but doesn't provide it
Quanah Gibson-Mount [Thu, 3 Nov 2011 23:04:25 +0000 (16:04 -0700)]
ITS#7078
Howard Chu [Thu, 6 Oct 2011 23:10:47 +0000 (16:10 -0700)]
Additional tweaks for double-threaded slapadd
Howard Chu [Wed, 5 Oct 2011 06:45:33 +0000 (23:45 -0700)]
Additional change for prev commit
Howard Chu [Wed, 5 Oct 2011 02:19:25 +0000 (19:19 -0700)]
Use a thread for LDIF parsing in slapadd -q
Quanah Gibson-Mount [Thu, 3 Nov 2011 23:01:12 +0000 (16:01 -0700)]
ITS#6737
Pierangelo Masarati [Tue, 7 Dec 2010 10:50:54 +0000 (10:50 +0000)]
move ctxcsn and schema check code in helpers; also apply to slapmodify (ITS#6737)
Quanah Gibson-Mount [Thu, 3 Nov 2011 22:08:21 +0000 (15:08 -0700)]
ITS#6437
Howard Chu [Fri, 9 Sep 2011 04:39:45 +0000 (21:39 -0700)]
Add Valgrind mempool support.
Compile with -DUSE_VALGRIND. It will attempt to keep Valgrind
informed of our memory allocations. Frees are trickier since they
may be deferred indefinitely.
Howard Chu [Thu, 23 Jun 2011 03:03:02 +0000 (20:03 -0700)]
Fix NO_THREADS typo
Hallvard Furuseth [Tue, 5 Jan 2010 23:31:14 +0000 (23:31 +0000)]
ITS#6437: Fix pool version misalignment: It aligned heads, not returned blocks
Hallvard Furuseth [Tue, 5 Jan 2010 23:14:14 +0000 (23:14 +0000)]
ITS#6437, move structs slab_object and slab_heap from slap.h to sl_malloc.c
Hallvard Furuseth [Tue, 5 Jan 2010 22:11:24 +0000 (22:11 +0000)]
ITS#6437, save space: Do not allocate the tail, except if size==0.
Store a tail only in freed blocks. (Alignment ensures there will
be room.) Put the freed mark in next block's head.
Hallvard Furuseth [Tue, 5 Jan 2010 19:40:28 +0000 (19:40 +0000)]
ITS#6437 cleanup (noop patch): Join SLAP_NO_SL_MALLOC code with identical
ctx==NULL code. Add enum No_sl_malloc to avoid #ifdef SLAP_NO_SL_MALLOC.
Hallvard Furuseth [Tue, 5 Jan 2010 19:25:28 +0000 (19:25 +0000)]
ITS#6437, follow up previous fixes:
Ensure Align >= sizeof(ber_len_t), to clarify the code's requirements.
Rearrange a slap_sl_malloc() branch to factor out ch_malloc fallbacks.
Fix range check (ptr+size >= endptr) -> (size >= endptr-ptr). Fix debug msg.
Hallvard Furuseth [Sat, 2 Jan 2010 23:42:31 +0000 (23:42 +0000)]
Add comments
Hallvard Furuseth [Sat, 2 Jan 2010 22:53:22 +0000 (22:53 +0000)]
ITS#6437, slap_sl_calloc():
Check count*size overflow. Omit slap_sl_malloc failure check, it cannot fail.
Quanah Gibson-Mount [Thu, 3 Nov 2011 22:03:48 +0000 (15:03 -0700)]
ITS#6437, failure/fallback/debug handling:
slap_sl_malloc could return failure. Exit instead, like the rest of sl_malloc.
Since we increase size, reduce it before Debug() and/or fallback to ch_malloc.
Debug() before exit() on failure. Tweak debug formats and a comment.
Conflicts:
servers/slapd/sl_malloc.c
Hallvard Furuseth [Sat, 2 Jan 2010 22:03:48 +0000 (22:03 +0000)]
ITS#6437 cleanup (noop change):
Move thread/nothread code to macros SET_MEMCTX()/GET_MEMCTX().
Rename some thread/memory contexts 'ctx' to thrctx/memctx to avoid confusion.
Hallvard Furuseth [Sat, 2 Jan 2010 21:39:46 +0000 (21:39 +0000)]
ITS#6437, fixes for stack implementation of slap_sl_realloc():
Preserve 2*int alignment on hosts where sizeof(int) == sizeof(ber_len_t).
If realloc of last block falls back to ch_malloc, free last block properly.
Fix range check (ptr + size < endptr) --> (size < endptr - ptr).
Hallvard Furuseth [Sat, 2 Jan 2010 21:05:11 +0000 (21:05 +0000)]
ITS#6437: Fix alignment after slap_sl_realloc().
Also use -Align instead of ~pad: Valid also for non-twos complement.
Hallvard Furuseth [Sat, 2 Jan 2010 18:42:33 +0000 (18:42 +0000)]
ITS#6437 cleanup (noop change): Simplify slap_sl_free() code reclaiming space
Hallvard Furuseth [Sat, 2 Jan 2010 18:27:51 +0000 (18:27 +0000)]
ITS#6437 cleanup (noop change): Compute pad,order_start at compile time
Hallvard Furuseth [Sat, 2 Jan 2010 18:05:56 +0000 (18:05 +0000)]
ITS#6437, slap_sl_mem_create():
Bugfix: Switching implementation stack<->pool on an old context could
crash or leak, it ran the wrong implementation's cleanup code.
Cleanup: Factor out identical stack/pool code. Call
slap_sl_mem_destroy(NULL,) instead of copying its pool code.
API change: slap_sl_mem_destroy(key=NULL,) gets a new meaning. Does
not affect current OpenLDAP code, it never passed NULL.