Kurt Zeilenga [Tue, 22 Dec 1998 19:08:27 +0000 (19:08 +0000)]
Update man pages based upon required use of ldap_memfree() instead of free()
replace 'malloced' with 'dynamically allocated'
State ldap_memfree(3) should be used instead of free()
Still need ldap_memfree(3) man page.
Kurt Zeilenga [Mon, 21 Dec 1998 23:08:06 +0000 (23:08 +0000)]
Still wasn't working exactly write on FreeBSD3. Need to
check for $OBJFORMAT and if not available set to output of
/usr/bin/objformat (if available). Otherwise assume not elf.
Kurt Zeilenga [Mon, 21 Dec 1998 17:45:47 +0000 (17:45 +0000)]
ITS#12 realBase was being freed before use. The offending free()
call has been removed and additional code added to properly free
realBase variable. Also fixed leaking of the matched variable.
Kurt Zeilenga [Mon, 21 Dec 1998 00:21:58 +0000 (00:21 +0000)]
LDAP C-API changes
struct friendly (Friendly) changed to ldap_friendly (LDAPFriendly)
ldap_friendly friend prefix changed to 'lf_' from 'f_'
removed mod_next field from LDAPMod (struct ldapmod)
modified slapd to use new LDAPModList (struct ldapmodlist) struct.
Added LDAPv3 result codes to ldap.h
Kurt Zeilenga [Sun, 20 Dec 1998 18:16:03 +0000 (18:16 +0000)]
Minor change to signal handler. Changed _shutdown flag to
type 'sig_atomic_t' and marked it volatile. Add detection to
configure to ensure sig_atomic_t and volatile are supported.
Kurt Zeilenga [Sat, 19 Dec 1998 01:27:20 +0000 (01:27 +0000)]
-lldap reentrantancy improvements from Bart Hartgers <A.Hartgers@phys.tue.nl>
including use of ctime_r, gethostby*_r, etc.. Also reworked ldap_sort_entries
to be reentrant.
Need to add code to properly set LDAP_API_FEATURE_THREAD_SAFE.
Kurt Zeilenga [Fri, 18 Dec 1998 21:54:06 +0000 (21:54 +0000)]
Fix ltconfig for FreeBSD 3 (elf detection was wrong).
Also fixed "empty translation unit" problem reported by
Randy Kunkee <kunkee@pluto.ops.NeoSoft.com>
Kurt Zeilenga [Fri, 18 Dec 1998 18:06:30 +0000 (18:06 +0000)]
Patch cache_delete_entry_internal() as suggested by
Daniel Carroll <dan@mesa7.mesa.colorado.edu>
This patch ensures we try deleting the entry from both the
dntree and the idtree.
Kurt Spanier [Fri, 18 Dec 1998 12:28:47 +0000 (12:28 +0000)]
Initialization of Datum structures to zero by memset() before first use.
If Berkeley DB version 2.x is used as a slapd back-end a not-initialized
Datum structure will raise an EINVAL error in (DB)->put().
Kurt Zeilenga [Mon, 14 Dec 1998 10:10:41 +0000 (10:10 +0000)]
Backout previous changes made by Dave as they excessive reformatting,
reindentation, and use of spaces instead of tab. Diffs will be reviewed
to identify and commit only functional changes as time permits.
Note: this is more my fault than Dave's. I should have provided
clearer guidance. Live and learn.
Kurt Zeilenga [Mon, 14 Dec 1998 04:25:55 +0000 (04:25 +0000)]
The result set of an EQUALITY search on a SYNTAX_BIN attribute may have been
incorrect because value_cmp() in value.c compares two attributes by
comparing the leading bytes of the two values (candidate and filter) up to
the smaller of the two lengths of the binary values. If the lengths of two
SYNTAX_BIN values differ, like SYNTAX_CES and SYNTAX_CIS, the EQUALITY
comparison should fail.
It also appears that value_ncmp() in value.c has no references anywhere in
the openLDAP source tree, save for a definition in proto-slap.h. The
following fix the bug and remove the dead code:
Kurt Zeilenga [Mon, 14 Dec 1998 04:21:04 +0000 (04:21 +0000)]
Make sure that four characters, rather than only the first three,
are being analysed in the transaltion to base64.
From: Lucio de Re <lucio@proxima.alt.za>
Kurt Zeilenga [Mon, 14 Dec 1998 00:23:22 +0000 (00:23 +0000)]
Update pthread checks:
Add plain -lpthread check.
Move *_r() checks so that they are checked for regradless
of thread libs used. Before they were only being checked
when using threads.
Fix yields test such that compile/link/odd runtime failures results in
'yes' instead of 'no'.
Kurt Zeilenga [Thu, 10 Dec 1998 19:20:15 +0000 (19:20 +0000)]
Change cache.c assert logic based on Hallvard's suggestion.
"Check that the entry has no writers before removing it from the
cache" sounds more sensible to me than "check that it *has* writers"
before removing it.
I've also changed the default cachesize for tests to 4. This
should help discover such problems.