Commit Graph

122 Commits

Author SHA1 Message Date
Kurt Zeilenga
63f5ec8f8b remove ;; typo 1999-02-05 17:39:38 +00:00
Kurt Spanier
d92b9d3c9e Introduction of a new Berkeley DB version 2 (!) specific backend.
BEWARE: the backend will be compiled.
        the backend will NOT be invoked, yet.
        the backend CANNOT be invoked, yet, because it is NOT yet
            integrated into the new initialization/startup environment
            of the slapd server.
1999-02-05 16:23:03 +00:00
Kurt Spanier
8107ec91f0 Fix braindead signal handling in linux kernal threads.
First version with waiting on int to become 0; should be implemented
  via conditional variable...
Fix of typos.
First anchor for new bdb2 backend.
1999-02-05 13:20:57 +00:00
Kurt Zeilenga
216049bd12 New Frontend->Backend Interface
Separates per backend type from per backend database initialization
	and startup.  Also supports per type / per backend shutdown.
New frontend startup/shutdown routines are also provided:
	slap_init() slap_startup() slap_shutdown() slap_destroy()
New frontend->backend startup/shutdown is managed by:
	backend_init() backend_startup() backend_shutdown backend_destroy
backend_init() now calls bi_init() to initial all function pointers
for the backend (excepting bi_init() which is now the only hardcoded
entry point).  New entry points are detailed in slap.h struct
backend_info.  backend_info is a per database type structure.
Besides the new startup/shutdown entry points, the new interface
also supports per backend type configuration options.  One could have:

	backend bdb2	(new Berkeley DB 2 backend)
	bdb2_home	/directory

	database bdb2
	...

	*** This code is fairly experimental ***
	*** Much cleanup and testing is still needed ***

see slap.h for details on struct backend_db and backend_info.
1999-02-05 09:03:47 +00:00
Will Ballantyne
44c5c21078 additional checks around derefAlias_r call:
check for alias problem
check for alias != entry
check for null entry in subsequent switch statement
1999-02-04 18:55:58 +00:00
Kurt Zeilenga
366701bdf7 Fix dbcache/entry lock deadlock. If dbcache lock is held, it's
okay to read and write LDBM specific fields (state, refcnt,
LRU.  The id field, though is read-only once set.
cache_find_entry_dn2id(), hence, does not require any entry locks.
cache_find_entry_id() must do a entry_rdwr_trylock() and back
off if busy.
Add new rdwr lock code with trylock() functionality.
Implement entry_rdwr_trylock().
1999-02-04 18:00:50 +00:00
Kurt Zeilenga
fc44d42c53 Backout startup/initialization changes in preparation for new
frontend->backend interface.
1999-02-04 17:28:49 +00:00
Kurt Zeilenga
c6882d4670 Remove rwlock diag messages. 1999-02-04 17:24:26 +00:00
Kurt Zeilenga
efce9cffb4 Use LDAP_ALIAS_DEREF_PROBLEM for all problem dereferencing aliases.
LDAP_ALIAS_PROBLEM is now only used if the alias itself is mangled.
1999-02-04 17:23:34 +00:00
Will Ballantyne
be9e28907e inserted additional checks for null values, comments
fixed deadlocks
1999-02-04 00:21:57 +00:00
Kurt Zeilenga
7903584911 Use idl_fetch_one() to fetch single id_blocks instead of separate
ldbm_cache_fetch() codes.
1999-02-03 17:14:01 +00:00
Kurt Zeilenga
ace9fb68c5 Fix ALLIDS bugs
Fix memory leaks
1999-02-03 04:45:08 +00:00
Kurt Zeilenga
16b3111ac7 Got that logic backwards... 1999-02-03 03:50:11 +00:00
Kurt Zeilenga
0c2ce661fb first arg of ldbm_db_errcall could be a const char * instead of char *. 1999-02-02 23:15:45 +00:00
Kurt Zeilenga
40ba9bc80b Don't try to free NULL idl. Did not cause a problem, though, as
idl_free() had insurance.
1999-02-02 22:46:21 +00:00
Kurt Zeilenga
92e71b459b Hack up a quick bandaid to make things compile. Startup/Initialization
needs work.
1999-02-02 19:26:00 +00:00
Kurt Spanier
442c7e7a7b Added startup/shutdown routine for slapd back-ends. 1999-02-02 08:03:03 +00:00
Kurt Zeilenga
433e8ad645 Add a few comments to help detail some of these functions. 1999-02-02 03:40:19 +00:00
Kurt Zeilenga
3e48937ac8 Fix idl_firstid to return NOID when !(nids > 1) instead of when (nids == 1)
Fix idl_nextid not to skip an ID if ID not in list.
Both of these should not occur if used correctly, but cheap insurance
against incorrect usage is welcomed.
1999-02-02 02:46:18 +00:00
Kurt Spanier
314f3f6ed7 Introduction of a backend startup/shutdown function to make backend-specific
initialization after reading slapd.conf, and before starting the daemon
1999-02-01 17:37:43 +00:00
Kurt Zeilenga
59a288abb5 Rewrite IDL block code to use simple array and accessor macros
instead of array with overlayed struct fields.  (which is inherently
non-portable).
Change maxid check in insert to be greater than or equal instead
of equal.
Use SAFEMEMCPY.
1999-01-31 07:55:53 +00:00
Kurt Zeilenga
1d70b89234 Make sure e_ndn is used! 1999-01-28 17:32:59 +00:00
Kurt Zeilenga
72ba4cfb71 Use -lldap_r instead of -lldap -lthread.
Likely broke things for non-posix threadings....

Update -lldap_r implementation to:
	remove attribute support
	hide thread detachment
	provide concurrency accessors
	provide initialization function
	fix gethostby{addr,name}_r codes (not coverred by HAVE_REENTRANT_FUNCTIONS)
Update servers/libraries to use ldap_pvt_thread_ calls.
Cleanup server codes (no #ifdef HAVE_PTHREAD_THIS or _THATs)!
Removed -llthread
1999-01-28 04:34:55 +00:00
Kurt Zeilenga
bd6dfe333d LDBM/back-ldbm locking reworked!
back-ldbm was using a cache specific lock to protect non-reentrant
	database routines from being reenterred.  This is inadequate.
	Also, reentrant database systems calls were serialized unnecessarily.

	Non-reentrant database calls must have a big_mutex.  Implemented
	this within -lldbm itself.  library requires ldbm_initialize()
	be called before any other ldbm call to initialize the big_mutex
	and to do any other db specific initialization (ie: such as
	required for DB2).

The dbc_mutex, dbc_cv, & dbc_readers fileds of dbcache are history.
The "reentrant_database" (REENTRANT_DATABASE) define is also axed.
1999-01-26 20:55:54 +00:00
Kurt Zeilenga
f1e934effb Hide allocate of reentrant_database cond varible and reader count
behind REENTRANT_DATABASE.  rename reentrant_database -> REENTRANT_DATABASE.
When REENTRANT_DATABASE is defined, the simple lock is replaced
with a simple reader/writer lock.  This functionality should be removed
from back-ldbm to libldbm.
1999-01-26 17:23:50 +00:00
Kurt Zeilenga
3c67248a65 cache_set_state( cache, e, 0) cannot occur before lock on parent/root
is released.  Some other thread may have cache locked but blocking
on parent.  Must give up parent lock before acquiring cache lock.
1999-01-26 05:22:58 +00:00
Kurt Zeilenga
4194c3ee85 If dn2id returns ID but id2entry returns NULL, log it.
Assume entry was deleted from underneath the dn2entry call,  hence
processing is same as if NOID had been returned.
1999-01-23 19:33:53 +00:00
Kurt Zeilenga
0f17fac37d Implement NEXTID chunking. Obtain IDs in chunks of size
SLAPD_NEXTID_CHUNK.  Code protects NEXTID file to ensure
its equal to or greater than nextid.  Updated on close
to actual nextid.  next_id_save() could be called periodically
if desired.  Default chunk size is 32.  Define to 1 to disable
chunking.
1999-01-22 02:26:19 +00:00
Kurt Zeilenga
5389e40ba5 initialize nextid with NOID not -1. 1999-01-22 01:53:39 +00:00
Kurt Zeilenga
2318ea0fd6 Don't return nextid on failed add operations for safety. 1999-01-22 01:51:53 +00:00
Kurt Zeilenga
5bfb3ac221 nextid cleanup in preparation to disable next_id_return(). 1999-01-22 01:40:39 +00:00
Kurt Zeilenga
ecdf9a354d Entry must be locked before adding it to the cache.
This removes a race condition upon the entry.
1999-01-22 00:27:50 +00:00
Kurt Zeilenga
6edeeeffd4 During cleanup always unlock root_mutex if rootlock is true
(don't make it conditional on p == NULL)
1999-01-21 22:20:03 +00:00
Kurt Zeilenga
f9b416ea7c ndn & strcasecmp cleanup 1999-01-20 03:05:35 +00:00
Kurt Zeilenga
8c501c28a2 Need to upper case the new_ndn. 1999-01-19 18:16:12 +00:00
Kurt Zeilenga
0e142b42f3 Update support for dn vs ndn. Build new dn from e->e_dn not frontend
provided dn (which is normalized and uppercased).   This preserved
the trailing portion of the dn.
1999-01-19 18:13:30 +00:00
Kurt Zeilenga
e2a15115b0 Update slap_conn to maintain client provided dn and bound dn.
Update slap_op to maintain dn and ndn (derived from conn->c_dn).
Update ldbm_back_bind to return actual bound dn (including rootdn)
	for use in slapd_conn.  Other backends use client dn.
Modify other codes to use ndn (normalized uppercase dn) most everywhere.
Aliasing, Suffixing and modrdn could use more work.
Applied suffixing to compare and modrdn.
1999-01-19 05:10:50 +00:00
Kurt Zeilenga
01a759971e Free that unclobberred dn. 1999-01-18 20:49:07 +00:00
Kurt Zeilenga
802325542b Don't clobber dn. 1999-01-18 20:34:17 +00:00
Kurt Zeilenga
c9242a123f e_ndn = dn_normalize_case( e_dn ) 1999-01-18 20:00:58 +00:00
Kurt Zeilenga
8009e591ce Protect return of the locked entry.
Note: we probably can replace the dn2entry_r with dn2id as only the id
is needed unless we're holding the lock for filtercandidates or something
1999-01-18 17:41:12 +00:00
Kurt Zeilenga
818bc57319 Minor cleanup of trace output. 1999-01-18 17:33:36 +00:00
Kurt Zeilenga
77b2e2cb25 Use Entry's e_ndn instead of recomputing it. 1999-01-17 20:12:00 +00:00
Kurt Zeilenga
3e3f0b8871 Could have no entry to return... check entry pointer before trying to
return cache entry.
1999-01-15 18:04:52 +00:00
Kurt Zeilenga
b39fdc0bcd Add normalized dn to Entry structure as field e_ndn. Entry
creation codes to provide this field.  Update cache_entrydn_cmp
to strcasecmp() the e_ndn instead of e_dn.  Note: strcasecmp()
is still used as e_ndn isn't in uppercase.  Maybe it should
be.  Did not update other codes to use e_ndn.  Hence, there
are lots of dn_normalize() calls that could be eliminated.
(The case determination of e_ndn should be made first).
1999-01-13 01:17:22 +00:00
Kurt Zeilenga
1d7704b28e Fix tmp file removal. 1999-01-08 19:51:15 +00:00
Kurt Zeilenga
b815f61e0d Wrap db2 mutex with -DHAVE_BERKELEY_DB2 (should be hidden in -lldbm) ITS#35
Unwrap ldbm_datum_init()
1999-01-07 03:28:08 +00:00
Kurt Zeilenga
0f76c52448 Modify Add/Delete/Modrdn operations to require write perms on
parent's "children" attribute.  Write lock parent to prevent
multiple clients making conflicting operations concurrently.
If parent doesn't exist (ie: is backend root), acquire a writer
lock (a simple mutex) on the "root."
Use -DSLAPD_CHILD_MODIFICATION_WITH_ENTRY_ACL to use the child's
"entry" acl for modrdn/delete operations.
1999-01-07 02:51:08 +00:00
Kurt Spanier
9c6d384b9a - Make install creates $(RUNDIR)/var for pid and args files of slapd.
- slapd's pid/args file names are based on the servers binary file names,
  providing for multiple servers beeing run on one host.
- slapd supports the -l command line parameter for selection of a
  syslog LOCAL user (-lLOCAL0 .. -lLOCAL7)
- db_appinit() is called during first ldbm_open() in DB 2.x to initialize
  DB debugging features (good to find bugs in the DB code :-)
- a patch for a non-initialized variable in DB's 2.x db_open is provided.
1999-01-05 15:40:58 +00:00
Kurt Zeilenga
c97837277a entry pointer 'p' must be initialized. Should fix ITS#31. 1999-01-03 22:55:43 +00:00