Adding upstream version 2:4.17.12+dfsg.upstream/2%4.17.12+dfsgupstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 1242 insertions, 0 deletions

diff --git a/lib/ldb/common/ldb_modules.c b/lib/ldb/common/ldb_modules.c
new file mode 100644
index 0000000..2d0494a
--- /dev/null
+++ b/lib/ldb/common/ldb_modules.c
@@ -0,0 +1,1242 @@
+/*
+   ldb database library
+
+   Copyright (C) Simo Sorce  2004-2008
+
+     ** NOTE! The following LGPL license applies to the ldb
+     ** library. This does NOT imply that all of Samba is released
+     ** under the LGPL
+
+   This library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 3 of the License, or (at your option) any later version.
+
+   This library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with this library; if not, see <http://www.gnu.org/licenses/>.
+*/
+
+/*
+ *  Name: ldb
+ *
+ *  Component: ldb modules core
+ *
+ *  Description: core modules routines
+ *
+ *  Author: Simo Sorce
+ */
+
+#include "ldb_private.h"
+#include "dlinklist.h"
+#include "system/dir.h"
+
+static char *ldb_modules_strdup_no_spaces(TALLOC_CTX *mem_ctx, const char *string)
+{
+	size_t i, len;
+	char *trimmed;
+
+	trimmed = talloc_strdup(mem_ctx, string);
+	if (!trimmed) {
+		return NULL;
+	}
+
+	len = strlen(trimmed);
+	for (i = 0; trimmed[i] != '\0'; i++) {
+		switch (trimmed[i]) {
+		case ' ':
+		case '\t':
+		case '\n':
+			memmove(&trimmed[i], &trimmed[i + 1], len -i -1);
+			break;
+		}
+	}
+
+	return trimmed;
+}
+
+
+/* modules are called in inverse order on the stack.
+   Lets place them as an admin would think the right order is.
+   Modules order is important */
+const char **ldb_modules_list_from_string(struct ldb_context *ldb, TALLOC_CTX *mem_ctx, const char *string)
+{
+	char **modules = NULL;
+	const char **m;
+	char *modstr, *p;
+	unsigned int i;
+
+	/* spaces not admitted */
+	modstr = ldb_modules_strdup_no_spaces(mem_ctx, string);
+	if ( ! modstr) {
+		ldb_debug(ldb, LDB_DEBUG_FATAL, "Out of Memory in ldb_modules_strdup_no_spaces()");
+		return NULL;
+	}
+
+	modules = talloc_realloc(mem_ctx, modules, char *, 2);
+	if ( ! modules ) {
+		ldb_debug(ldb, LDB_DEBUG_FATAL, "Out of Memory in ldb_modules_list_from_string()");
+		talloc_free(modstr);
+		return NULL;
+	}
+	talloc_steal(modules, modstr);
+
+	if (modstr[0] == '\0') {
+		modules[0] = NULL;
+		m = discard_const_p(const char *, modules);
+		return m;
+	}
+
+	i = 0;
+	/* The str*r*chr walks backwards:  This is how we get the inverse order mentioned above */
+	while ((p = strrchr(modstr, ',')) != NULL) {
+		*p = '\0';
+		p++;
+		modules[i] = p;
+
+		i++;
+		modules = talloc_realloc(mem_ctx, modules, char *, i + 2);
+		if ( ! modules ) {
+			ldb_debug(ldb, LDB_DEBUG_FATAL, "Out of Memory in ldb_modules_list_from_string()");
+			return NULL;
+		}
+
+	}
+	modules[i] = modstr;
+
+	modules[i + 1] = NULL;
+
+	m = discard_const_p(const char *, modules);
+
+	return m;
+}
+
+static struct backends_list_entry {
+	struct ldb_backend_ops *ops;
+	struct backends_list_entry *prev, *next;
+} *ldb_backends = NULL;
+
+static struct ops_list_entry {
+	const struct ldb_module_ops *ops;
+	struct ops_list_entry *next;
+} *registered_modules = NULL;
+
+static struct backends_list_entry *ldb_find_backend(const char *url_prefix)
+{
+	struct backends_list_entry *backend;
+
+	for (backend = ldb_backends; backend; backend = backend->next) {
+		if (strcmp(backend->ops->name, url_prefix) == 0) {
+			return backend;
+		}
+	}
+
+	return NULL;
+}
+
+/*
+  register a new ldb backend
+
+  if override is true, then override any existing backend for this prefix
+*/
+int ldb_register_backend(const char *url_prefix, ldb_connect_fn connectfn, bool override)
+{
+	struct backends_list_entry *be;
+
+	be = ldb_find_backend(url_prefix);
+	if (be) {
+		if (!override) {
+			return LDB_SUCCESS;
+		}
+	} else {
+		be = talloc_zero(ldb_backends, struct backends_list_entry);
+		if (!be) {
+			return LDB_ERR_OPERATIONS_ERROR;
+		}
+		be->ops = talloc_zero(be, struct ldb_backend_ops);
+		if (!be->ops) {
+			talloc_free(be);
+			return LDB_ERR_OPERATIONS_ERROR;
+		}
+		DLIST_ADD_END(ldb_backends, be);
+	}
+
+	be->ops->name = url_prefix;
+	be->ops->connect_fn = connectfn;
+
+	return LDB_SUCCESS;
+}
+
+/*
+   Return the ldb module form of a database.
+   The URL looks something like this:
+     tdb://PATH
+     ldb://PATH
+     mdb://PATH
+     ldapi://PATH
+     PATH          (unadorned PATH defaults to tdb://)
+
+   for a complete list of backends (including possibly unmaintained ones) grep
+   for calls to ldb_register_backend().
+
+   the options are passed uninterpreted to the backend, and are
+   backend specific.
+
+   This allows modules to get at only the backend module, for example where a
+   module may wish to direct certain requests at a particular backend.
+*/
+int ldb_module_connect_backend(struct ldb_context *ldb,
+			       const char *url,
+			       const char *options[],
+			       struct ldb_module **backend_module)
+{
+	int ret;
+	char *backend;
+	struct backends_list_entry *be;
+	char *colon = NULL;
+
+	colon = strchr(url, ':');
+	if (colon != NULL) {
+		backend = talloc_strndup(ldb, url, colon-url);
+	} else {
+		/* Default to tdb */
+		backend = talloc_strdup(ldb, "tdb");
+	}
+	if (backend == NULL) {
+		return ldb_oom(ldb);
+	}
+
+	be = ldb_find_backend(backend);
+
+	talloc_free(backend);
+
+	if (be == NULL) {
+		ldb_debug(ldb, LDB_DEBUG_FATAL,
+			  "Unable to find backend for '%s' - do you need to set LDB_MODULES_PATH?", url);
+		return LDB_ERR_OTHER;
+	}
+
+	ret = be->ops->connect_fn(ldb, url, ldb->flags, options, backend_module);
+
+	if (ret != LDB_SUCCESS) {
+		ldb_debug(ldb, LDB_DEBUG_ERROR,
+			  "Failed to connect to '%s' with backend '%s': %s", url, be->ops->name, ldb_errstring(ldb));
+		return ret;
+	}
+	return ret;
+}
+
+static struct ldb_hooks {
+	struct ldb_hooks *next, *prev;
+	ldb_hook_fn hook_fn;
+} *ldb_hooks;
+
+/*
+  register a ldb hook function
+ */
+int ldb_register_hook(ldb_hook_fn hook_fn)
+{
+	struct ldb_hooks *lc;
+	lc = talloc_zero(ldb_hooks, struct ldb_hooks);
+	if (lc == NULL) {
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+	lc->hook_fn = hook_fn;
+	DLIST_ADD_END(ldb_hooks, lc);
+	return LDB_SUCCESS;
+}
+
+/*
+  call ldb hooks of a given type
+ */
+int ldb_modules_hook(struct ldb_context *ldb, enum ldb_module_hook_type t)
+{
+	struct ldb_hooks *lc;
+	for (lc = ldb_hooks; lc; lc=lc->next) {
+		int ret = lc->hook_fn(ldb, t);
+		if (ret != LDB_SUCCESS) {
+			return ret;
+		}
+	}
+	return LDB_SUCCESS;
+}
+
+
+static const struct ldb_module_ops *ldb_find_module_ops(const char *name)
+{
+	struct ops_list_entry *e;
+
+	for (e = registered_modules; e; e = e->next) {
+ 		if (strcmp(e->ops->name, name) == 0)
+			return e->ops;
+	}
+
+	return NULL;
+}
+
+
+int ldb_register_module(const struct ldb_module_ops *ops)
+{
+	struct ops_list_entry *entry;
+
+	if (ldb_find_module_ops(ops->name) != NULL)
+		return LDB_ERR_ENTRY_ALREADY_EXISTS;
+
+	/*
+	 * ldb modules are not (yet) unloaded and
+	 * are only loaded once (the above check
+	 * makes sure of this). Allocate off the NULL
+	 * context. We never want this to be freed
+	 * until process shutdown. If eventually we
+	 * want to unload ldb modules we can add a
+	 * deregister function that walks and
+	 * frees the list.
+	 */
+	entry = talloc(NULL, struct ops_list_entry);
+	if (entry == NULL) {
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+
+	entry->ops = ops;
+	entry->next = registered_modules;
+	registered_modules = entry;
+
+	return LDB_SUCCESS;
+}
+
+/*
+  load a list of modules
+ */
+int ldb_module_load_list(struct ldb_context *ldb, const char **module_list,
+			 struct ldb_module *backend, struct ldb_module **out)
+{
+	struct ldb_module *module;
+	unsigned int i;
+
+	module = backend;
+
+	for (i = 0; module_list && module_list[i] != NULL; i++) {
+		struct ldb_module *current;
+		const struct ldb_module_ops *ops;
+
+		if (strcmp(module_list[i], "") == 0) {
+			continue;
+		}
+
+		ops = ldb_find_module_ops(module_list[i]);
+
+		if (ops == NULL) {
+			ldb_debug(ldb, LDB_DEBUG_FATAL, "WARNING: Module [%s] not found - do you need to set LDB_MODULES_PATH?",
+				  module_list[i]);
+			return LDB_ERR_OPERATIONS_ERROR;
+		}
+
+		current = talloc_zero(ldb, struct ldb_module);
+		if (current == NULL) {
+			return LDB_ERR_OPERATIONS_ERROR;
+		}
+		talloc_set_name(current, "ldb_module: %s", module_list[i]);
+
+		current->ldb = ldb;
+		current->ops = ops;
+
+		DLIST_ADD(module, current);
+	}
+	*out = module;
+	return LDB_SUCCESS;
+}
+
+/*
+  initialise a chain of modules
+ */
+int ldb_module_init_chain(struct ldb_context *ldb, struct ldb_module *module)
+{
+	while (module && module->ops->init_context == NULL)
+		module = module->next;
+
+	/* init is different in that it is not an error if modules
+	 * do not require initialization */
+
+	if (module) {
+		int ret = module->ops->init_context(module);
+		if (ret != LDB_SUCCESS) {
+			ldb_debug(ldb, LDB_DEBUG_FATAL, "module %s initialization failed : %s",
+				  module->ops->name, ldb_strerror(ret));
+			return ret;
+		}
+	}
+
+	return LDB_SUCCESS;
+}
+
+int ldb_load_modules(struct ldb_context *ldb, const char *options[])
+{
+	const char *modules_string;
+	const char **modules = NULL;
+	int ret;
+	TALLOC_CTX *mem_ctx = talloc_new(ldb);
+	if (!mem_ctx) {
+		return ldb_oom(ldb);
+	}
+
+	/* find out which modules we are requested to activate */
+
+	/* check if we have a custom module list passd as ldb option */
+	if (options) {
+		modules_string = ldb_options_find(ldb, options, "modules");
+		if (modules_string) {
+			modules = ldb_modules_list_from_string(ldb, mem_ctx, modules_string);
+		}
+	}
+
+	/* if not overloaded by options and the backend is not ldap try to load the modules list from ldb */
+	if ((modules == NULL) && (strcmp("ldap", ldb->modules->ops->name) != 0)) {
+		const char * const attrs[] = { "@LIST" , NULL};
+		struct ldb_result *res = NULL;
+		struct ldb_dn *mods_dn;
+
+		mods_dn = ldb_dn_new(mem_ctx, ldb, "@MODULES");
+		if (mods_dn == NULL) {
+			talloc_free(mem_ctx);
+			return ldb_oom(ldb);
+		}
+
+		ret = ldb_search(ldb, mods_dn, &res, mods_dn, LDB_SCOPE_BASE, attrs, "@LIST=*");
+
+		if (ret == LDB_ERR_NO_SUCH_OBJECT) {
+			ldb_debug(ldb, LDB_DEBUG_TRACE, "no modules required by the db");
+		} else if (ret != LDB_SUCCESS) {
+			ldb_debug(ldb, LDB_DEBUG_FATAL, "ldb error (%s) occurred searching for modules, bailing out", ldb_errstring(ldb));
+			talloc_free(mem_ctx);
+			return ret;
+		} else {
+			const char *module_list;
+			if (res->count == 0) {
+				ldb_debug(ldb, LDB_DEBUG_TRACE, "no modules required by the db");
+			} else if (res->count > 1) {
+				ldb_debug(ldb, LDB_DEBUG_FATAL, "Too many records found (%u), bailing out", res->count);
+				talloc_free(mem_ctx);
+				return LDB_ERR_OPERATIONS_ERROR;
+			} else {
+				module_list = ldb_msg_find_attr_as_string(res->msgs[0], "@LIST", NULL);
+				if (!module_list) {
+					ldb_debug(ldb, LDB_DEBUG_TRACE, "no modules required by the db");
+				}
+				modules = ldb_modules_list_from_string(ldb, mem_ctx,
+							       module_list);
+			}
+		}
+
+		talloc_free(mods_dn);
+	}
+
+	if (modules != NULL) {
+		ret = ldb_module_load_list(ldb, modules, ldb->modules, &ldb->modules);
+		if (ret != LDB_SUCCESS) {
+			talloc_free(mem_ctx);
+			return ret;
+		}
+	} else {
+		ldb_debug(ldb, LDB_DEBUG_TRACE, "No modules specified for this database");
+	}
+
+	ret = ldb_module_init_chain(ldb, ldb->modules);
+	talloc_free(mem_ctx);
+	return ret;
+}
+
+/*
+  by using this we allow ldb modules to only implement the functions they care about,
+  which makes writing a module simpler, and makes it more likely to keep working
+  when ldb is extended
+*/
+#define FIND_OP_NOERR(module, op) do { \
+	module = module->next; \
+	while (module && module->ops->op == NULL) module = module->next; \
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) { \
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE, "ldb_trace_next_request: (%s)->" #op, \
+			  module->ops->name);				\
+	}								\
+} while (0)
+
+#define FIND_OP(module, op) do { \
+	struct ldb_context *ldb = module->ldb; \
+	FIND_OP_NOERR(module, op); \
+	if (module == NULL) { \
+		ldb_asprintf_errstring(ldb, "Unable to find backend operation for " #op ); \
+		return LDB_ERR_OPERATIONS_ERROR;	\
+	}						\
+} while (0)
+
+
+struct ldb_module *ldb_module_new(TALLOC_CTX *memctx,
+				  struct ldb_context *ldb,
+				  const char *module_name,
+				  const struct ldb_module_ops *ops)
+{
+	struct ldb_module *module;
+
+	module = talloc(memctx, struct ldb_module);
+	if (!module) {
+		ldb_oom(ldb);
+		return NULL;
+	}
+	talloc_set_name_const(module, module_name);
+	module->ldb = ldb;
+	module->prev = module->next = NULL;
+	module->ops = ops;
+
+	return module;
+}
+
+const char * ldb_module_get_name(struct ldb_module *module)
+{
+	return module->ops->name;
+}
+
+struct ldb_context *ldb_module_get_ctx(struct ldb_module *module)
+{
+	return module->ldb;
+}
+
+const struct ldb_module_ops *ldb_module_get_ops(struct ldb_module *module)
+{
+	return module->ops;
+}
+
+void *ldb_module_get_private(struct ldb_module *module)
+{
+	return module->private_data;
+}
+
+void ldb_module_set_private(struct ldb_module *module, void *private_data)
+{
+	module->private_data = private_data;
+}
+
+/*
+   helper functions to call the next module in chain
+*/
+
+int ldb_next_request(struct ldb_module *module, struct ldb_request *request)
+{
+	int ret;
+
+	if (request->callback == NULL) {
+		ldb_set_errstring(module->ldb, "Requests MUST define callbacks");
+		return LDB_ERR_UNWILLING_TO_PERFORM;
+	}
+
+	request->handle->nesting++;
+
+	switch (request->operation) {
+	case LDB_SEARCH:
+		FIND_OP(module, search);
+		ret = module->ops->search(module, request);
+		break;
+	case LDB_ADD:
+		FIND_OP(module, add);
+		ret = module->ops->add(module, request);
+		break;
+	case LDB_MODIFY:
+		FIND_OP(module, modify);
+		ret = module->ops->modify(module, request);
+		break;
+	case LDB_DELETE:
+		FIND_OP(module, del);
+		ret = module->ops->del(module, request);
+		break;
+	case LDB_RENAME:
+		FIND_OP(module, rename);
+		ret = module->ops->rename(module, request);
+		break;
+	case LDB_EXTENDED:
+		FIND_OP(module, extended);
+		ret = module->ops->extended(module, request);
+		break;
+	default:
+		FIND_OP(module, request);
+		ret = module->ops->request(module, request);
+		break;
+	}
+
+	request->handle->nesting--;
+
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		const char *op;
+		switch (request->operation) {
+		case LDB_SEARCH:
+			op = "LDB_SEARCH";
+			break;
+		case LDB_ADD:
+			op = "LDB_ADD";
+			break;
+		case LDB_MODIFY:
+			op = "LDB_MODIFY";
+			break;
+		case LDB_DELETE:
+			op = "LDB_DELETE";
+			break;
+		case LDB_RENAME:
+			op = "LDB_RENAME";
+			break;
+		case LDB_EXTENDED:
+			op = "LDB_EXTENDED";
+			break;
+		default:
+			op = "request";
+			break;
+		}
+
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb, "error in module %s: %s during %s (%d)", module->ops->name, ldb_strerror(ret), op, ret);
+	}
+
+	if (!(request->handle->flags & LDB_HANDLE_FLAG_DONE_CALLED)) {
+		/* It is _extremely_ common that a module returns a
+		 * failure without calling ldb_module_done(), but that
+		 * guarantees we will end up hanging in
+		 * ldb_wait(). This fixes it without having to rewrite
+		 * all our modules, and leaves us one less sharp
+		 * corner for module developers to cut themselves on
+		 */
+		ret = ldb_module_done(request, NULL, NULL, ret);
+	}
+	return ret;
+}
+
+int ldb_next_init(struct ldb_module *module)
+{
+	module = module->next;
+
+	return ldb_module_init_chain(module->ldb, module);
+}
+
+int ldb_next_start_trans(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP(module, start_transaction);
+	ret = module->ops->start_transaction(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb, "start_trans error in module %s: %s (%d)", module->ops->name, ldb_strerror(ret), ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) { 
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE, "ldb_next_start_trans error: %s", 
+			  ldb_errstring(module->ldb));				
+	}
+	return ret;
+}
+
+int ldb_next_end_trans(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP(module, end_transaction);
+	ret = module->ops->end_transaction(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb, "end_trans error in module %s: %s (%d)", module->ops->name, ldb_strerror(ret), ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) { 
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE, "ldb_next_end_trans error: %s", 
+			  ldb_errstring(module->ldb));				
+	}
+	return ret;
+}
+
+int ldb_next_read_lock(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP(module, read_lock);
+	ret = module->ops->read_lock(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb,
+				       "read_lock error in module %s: %s (%d)",
+				       module->ops->name, ldb_strerror(ret),
+				       ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) {
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE,
+			  "ldb_next_read_lock error: %s",
+			  ldb_errstring(module->ldb));
+	}
+	return ret;
+}
+
+int ldb_next_read_unlock(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP(module, read_unlock);
+	ret = module->ops->read_unlock(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb,
+				       "read_unlock error in module %s: %s (%d)",
+				       module->ops->name, ldb_strerror(ret),
+				       ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) {
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE,
+			  "ldb_next_read_unlock error: %s",
+			  ldb_errstring(module->ldb));
+	}
+	return ret;
+}
+
+int ldb_next_prepare_commit(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP_NOERR(module, prepare_commit);
+	if (module == NULL) {
+		/* we are allowed to have no prepare commit in
+		   backends */
+		return LDB_SUCCESS;
+	}
+	ret = module->ops->prepare_commit(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb, "prepare_commit error in module %s: %s (%d)", module->ops->name, ldb_strerror(ret), ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) { 
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE, "ldb_next_prepare_commit error: %s", 
+			  ldb_errstring(module->ldb));				
+	}
+	return ret;
+}
+
+int ldb_next_del_trans(struct ldb_module *module)
+{
+	int ret;
+	FIND_OP(module, del_transaction);
+	ret = module->ops->del_transaction(module);
+	if (ret == LDB_SUCCESS) {
+		return ret;
+	}
+	if (!ldb_errstring(module->ldb)) {
+		/* Set a default error string, to place the blame somewhere */
+		ldb_asprintf_errstring(module->ldb, "del_trans error in module %s: %s (%d)", module->ops->name, ldb_strerror(ret), ret);
+	}
+	if ((module && module->ldb->flags & LDB_FLG_ENABLE_TRACING)) { 
+		ldb_debug(module->ldb, LDB_DEBUG_TRACE, "ldb_next_del_trans error: %s", 
+			  ldb_errstring(module->ldb));				
+	}
+	return ret;
+}
+
+/* calls the request callback to send an entry
+ *
+ * params:
+ *      req: the original request passed to your module
+ *      msg: reply message (must be a talloc pointer, and it will be stolen
+ *           on the ldb_reply that is sent to the callback)
+ * 	ctrls: controls to send in the reply  (must be a talloc pointer, and it will be stolen
+ *           on the ldb_reply that is sent to the callback)
+ */
+
+int ldb_module_send_entry(struct ldb_request *req,
+			  struct ldb_message *msg,
+			  struct ldb_control **ctrls)
+{
+	struct ldb_reply *ares;
+
+	ares = talloc_zero(req, struct ldb_reply);
+	if (!ares) {
+		ldb_oom(req->handle->ldb);
+		req->callback(req, NULL);
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+	ares->type = LDB_REPLY_ENTRY;
+	ares->message = talloc_steal(ares, msg);
+	ares->controls = talloc_steal(ares, ctrls);
+	ares->error = LDB_SUCCESS;
+
+	if ((req->handle->ldb->flags & LDB_FLG_ENABLE_TRACING) &&
+	    req->handle->nesting == 0) {
+		char *s;
+		struct ldb_ldif ldif;
+		
+		ldif.changetype = LDB_CHANGETYPE_NONE;
+		ldif.msg = discard_const_p(struct ldb_message, msg);
+
+		ldb_debug_add(req->handle->ldb, "ldb_trace_response: ENTRY\n");
+
+		/* 
+		 * The choice to call
+		 * ldb_ldif_write_redacted_trace_string() is CRITICAL
+		 * for security.  It ensures that we do not output
+		 * passwords into debug logs 
+		 */
+
+		s = ldb_ldif_write_redacted_trace_string(req->handle->ldb, msg, &ldif);
+		ldb_debug_add(req->handle->ldb, "%s\n", s);
+		talloc_free(s);
+		ldb_debug_end(req->handle->ldb, LDB_DEBUG_TRACE);
+	}
+
+	return req->callback(req, ares);
+}
+
+/* calls the request callback to send an referrals
+ *
+ * params:
+ *      req: the original request passed to your module
+ *      ref: referral string (must be a talloc pointer, steal)
+ */
+
+int ldb_module_send_referral(struct ldb_request *req,
+					   char *ref)
+{
+	struct ldb_reply *ares;
+
+	ares = talloc_zero(req, struct ldb_reply);
+	if (!ares) {
+		ldb_oom(req->handle->ldb);
+		req->callback(req, NULL);
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+	ares->type = LDB_REPLY_REFERRAL;
+	ares->referral = talloc_steal(ares, ref);
+	ares->error = LDB_SUCCESS;
+
+	if ((req->handle->ldb->flags & LDB_FLG_ENABLE_TRACING) &&
+	    req->handle->nesting == 0) {
+		ldb_debug_add(req->handle->ldb, "ldb_trace_response: REFERRAL\n");
+		ldb_debug_add(req->handle->ldb, "ref: %s\n", ref);
+		ldb_debug_end(req->handle->ldb, LDB_DEBUG_TRACE);
+	}
+
+	return req->callback(req, ares);
+}
+
+/* calls the original request callback
+ *
+ * params:
+ * 	req:   the original request passed to your module
+ * 	ctrls: controls to send in the reply (must be a talloc pointer, steal)
+ * 	response: results for extended request (steal)
+ * 	error: LDB_SUCCESS for a successful return
+ * 	       any other ldb error otherwise
+ */
+int ldb_module_done(struct ldb_request *req,
+		    struct ldb_control **ctrls,
+		    struct ldb_extended *response,
+		    int error)
+{
+	struct ldb_reply *ares;
+
+	ares = talloc_zero(req, struct ldb_reply);
+	if (!ares) {
+		ldb_oom(req->handle->ldb);
+		req->callback(req, NULL);
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+	ares->type = LDB_REPLY_DONE;
+	ares->controls = talloc_steal(ares, ctrls);
+	ares->response = talloc_steal(ares, response);
+	ares->error = error;
+
+	req->handle->flags |= LDB_HANDLE_FLAG_DONE_CALLED;
+
+	if ((req->handle->ldb->flags & LDB_FLG_ENABLE_TRACING) &&
+	    req->handle->nesting == 0) {
+		ldb_debug_add(req->handle->ldb, "ldb_trace_response: DONE\n");
+		ldb_debug_add(req->handle->ldb, "error: %d\n", error);
+		if (ldb_errstring(req->handle->ldb)) {
+			ldb_debug_add(req->handle->ldb, "msg: %s\n",
+				  ldb_errstring(req->handle->ldb));
+		}
+		ldb_debug_end(req->handle->ldb, LDB_DEBUG_TRACE);
+	}
+
+	return req->callback(req, ares);
+}
+
+/* to be used *only* in modules init functions.
+ * this function is synchronous and will register
+ * the requested OID in the rootdse module if present
+ * otherwise it will return an error */
+int ldb_mod_register_control(struct ldb_module *module, const char *oid)
+{
+	struct ldb_request *req;
+	int ret;
+
+	req = talloc_zero(module, struct ldb_request);
+	if (req == NULL) {
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+
+	req->operation = LDB_REQ_REGISTER_CONTROL;
+	req->op.reg_control.oid = oid;
+	req->callback = ldb_op_default_callback;
+
+	ldb_set_timeout(module->ldb, req, 0);
+
+	req->handle = ldb_handle_new(req, module->ldb);
+	if (req->handle == NULL) {
+		return LDB_ERR_OPERATIONS_ERROR;
+	}
+
+	ret = ldb_request(module->ldb, req);
+	if (ret == LDB_SUCCESS) {
+		ret = ldb_wait(req->handle, LDB_WAIT_ALL);
+	}
+	talloc_free(req);
+
+	return ret;
+}
+
+static int ldb_modules_load_dir(const char *modules_dir, const char *version);
+
+
+/*
+  load one module. A static list of loaded module inode numbers is
+  used to prevent a module being loaded twice
+
+  dlopen() is used on the module, and dlsym() is then used to look for
+  a ldb_init_module() function. If present, that function is called
+  with the ldb version number as an argument.
+
+  The ldb_init_module() function will typically call
+  ldb_register_module() and ldb_register_backend() to register a
+  module or backend, but it may also be used to register command line
+  handling functions, ldif handlers or any other local
+  modififications.
+
+  The ldb_init_module() function does not get a ldb_context passed in,
+  as modules will be used for multiple ldb context handles. The call
+  from the first ldb_init() is just a convenient way to ensure it is
+  called early enough.
+ */
+static int ldb_modules_load_path(const char *path, const char *version)
+{
+	void *handle;
+	int (*init_fn)(const char *);
+	int ret;
+	struct stat st;
+	static struct loaded {
+		struct loaded *next, *prev;
+		ino_t st_ino;
+		dev_t st_dev;
+	} *loaded;
+	struct loaded *le;
+	int dlopen_flags;
+
+#ifdef RTLD_DEEPBIND
+	bool deepbind_enabled = (getenv("LDB_MODULES_DISABLE_DEEPBIND") == NULL);
+#endif
+
+	ret = stat(path, &st);
+	if (ret != 0) {
+		fprintf(stderr, "ldb: unable to stat module %s : %s\n", path, strerror(errno));
+		return LDB_ERR_UNAVAILABLE;
+	}
+
+	for (le=loaded; le; le=le->next) {
+		if (le->st_ino == st.st_ino &&
+		    le->st_dev == st.st_dev) {
+			/* its already loaded */
+			return LDB_SUCCESS;
+		}
+	}
+
+	le = talloc(loaded, struct loaded);
+	if (le == NULL) {
+		fprintf(stderr, "ldb: unable to allocated loaded entry\n");
+		return LDB_ERR_UNAVAILABLE;
+	}
+
+	le->st_ino = st.st_ino;
+	le->st_dev = st.st_dev;
+
+	DLIST_ADD_END(loaded, le);
+
+	/* if it is a directory, recurse */
+	if (S_ISDIR(st.st_mode)) {
+		return ldb_modules_load_dir(path, version);
+	}
+
+	dlopen_flags = RTLD_NOW;
+#ifdef RTLD_DEEPBIND
+	/*
+	 * use deepbind if possible, to avoid issues with different
+	 * system library variants, for example ldb modules may be linked
+	 * against Heimdal while the application may use MIT kerberos.
+	 *
+	 * See the dlopen manpage for details.
+	 *
+	 * One typical user is the bind_dlz module of Samba,
+	 * but symbol versioning might be enough...
+	 *
+	 * We need a way to disable this in order to allow the
+	 * ldb_*ldap modules to work with a preloaded socket wrapper.
+	 *
+	 * So in future we may remove this completely
+	 * or at least invert the default behavior.
+	*/
+	if (deepbind_enabled) {
+		dlopen_flags |= RTLD_DEEPBIND;
+	}
+#endif
+
+	handle = dlopen(path, dlopen_flags);
+	if (handle == NULL) {
+		fprintf(stderr, "ldb: unable to dlopen %s : %s\n", path, dlerror());
+		return LDB_SUCCESS;
+	}
+
+	init_fn = dlsym(handle, "ldb_init_module");
+	if (init_fn == NULL) {
+		/* ignore it, it could be an old-style
+		 * module. Once we've converted all modules we
+		 * could consider this an error */
+		dlclose(handle);
+		return LDB_SUCCESS;
+	}
+
+	ret = init_fn(version);
+	if (ret == LDB_ERR_ENTRY_ALREADY_EXISTS) {
+		/* the module is already registered - ignore this, as
+		 * it can happen if LDB_MODULES_PATH points at both
+		 * the build and install directory
+		 */
+		ret = LDB_SUCCESS;
+	}
+	return ret;
+}
+
+static int qsort_string(const char **s1, const char **s2)
+{
+	return strcmp(*s1, *s2);
+}
+
+
+/*
+  load all modules from the given ldb modules directory. This is run once
+  during the first ldb_init() call.
+
+  Modules are loaded in alphabetical order to ensure that any module
+  load ordering dependencies are reproducible. Modules should avoid
+  relying on load order
+ */
+static int ldb_modules_load_dir(const char *modules_dir, const char *version)
+{
+	DIR *dir;
+	struct dirent *de;
+	const char **modlist = NULL;
+	TALLOC_CTX *tmp_ctx = talloc_new(NULL);
+	unsigned i, num_modules = 0;
+
+	dir = opendir(modules_dir);
+	if (dir == NULL) {
+		if (errno == ENOENT) {
+			talloc_free(tmp_ctx);
+			/* we don't have any modules */
+			return LDB_SUCCESS;
+		}
+		talloc_free(tmp_ctx);
+		fprintf(stderr, "ldb: unable to open modules directory '%s' - %s\n",
+			modules_dir, strerror(errno));
+		return LDB_ERR_UNAVAILABLE;
+	}
+
+
+	while ((de = readdir(dir))) {
+		if (ISDOT(de->d_name) || ISDOTDOT(de->d_name))
+			continue;
+
+		modlist = talloc_realloc(tmp_ctx, modlist, const char *, num_modules+1);
+		if (modlist == NULL) {
+			talloc_free(tmp_ctx);
+			closedir(dir);
+			fprintf(stderr, "ldb: unable to allocate modules list\n");
+			return LDB_ERR_UNAVAILABLE;
+		}
+		modlist[num_modules] = talloc_asprintf(modlist, "%s/%s", modules_dir, de->d_name);
+		if (modlist[num_modules] == NULL) {
+			talloc_free(tmp_ctx);
+			closedir(dir);
+			fprintf(stderr, "ldb: unable to allocate module list entry\n");
+			return LDB_ERR_UNAVAILABLE;
+		}
+		num_modules++;
+	}
+
+	closedir(dir);
+
+	/* sort the directory, so we get consistent load ordering */
+	TYPESAFE_QSORT(modlist, num_modules, qsort_string);
+
+	for (i=0; i<num_modules; i++) {
+		int ret = ldb_modules_load_path(modlist[i], version);
+		if (ret != LDB_SUCCESS) {
+			fprintf(stderr, "ldb: failed to initialise module %s : %s\n",
+				modlist[i], ldb_strerror(ret));
+			talloc_free(tmp_ctx);
+			return ret;
+		}
+	}
+
+	talloc_free(tmp_ctx);
+
+	return LDB_SUCCESS;
+}
+
+/* 
+   load any additional modules from the given directory 
+*/
+void ldb_set_modules_dir(struct ldb_context *ldb, const char *path)
+{
+	int ret = ldb_modules_load_dir(path, LDB_VERSION);
+	if (ret != LDB_SUCCESS) {
+		ldb_asprintf_errstring(ldb, "Failed to load modules from: %s\n", path);
+	}
+}
+
+
+/*
+  load all modules static (builtin) modules
+ */
+static int ldb_modules_load_static(const char *version)
+{
+	static bool initialised;
+#define _MODULE_PROTO(init) extern int init(const char *);
+	STATIC_ldb_MODULES_PROTO;
+	const ldb_module_init_fn static_init_functions[] = { STATIC_ldb_MODULES };
+	unsigned i;
+
+	if (initialised) {
+		return LDB_SUCCESS;
+	}
+	initialised = true;
+
+	for (i=0; static_init_functions[i]; i++) {
+		int ret = static_init_functions[i](version);
+		if (ret != LDB_SUCCESS) {
+			return ret;
+		}
+	}
+	return LDB_SUCCESS;
+}
+
+/*
+  load all modules from the given ldb modules path, colon
+  separated.
+
+  modules are loaded recursively for all subdirectories in the paths
+ */
+int ldb_modules_load(const char *modules_path, const char *version)
+{
+	char *tok, *path, *tok_ptr=NULL;
+	int ret;
+
+	ret = ldb_modules_load_static(version);
+	if (ret != LDB_SUCCESS) {
+		return ret;
+	}
+
+	path = talloc_strdup(NULL, modules_path);
+	if (path == NULL) {
+		fprintf(stderr, "ldb: failed to allocate modules_path\n");
+		return LDB_ERR_UNAVAILABLE;
+	}
+
+	for (tok=strtok_r(path, ":", &tok_ptr);
+	     tok;
+	     tok=strtok_r(NULL, ":", &tok_ptr)) {
+		ret = ldb_modules_load_path(tok, version);
+		if (ret != LDB_SUCCESS) {
+			talloc_free(path);
+			return ret;
+		}
+	}
+	talloc_free(path);
+
+	return LDB_SUCCESS;
+}
+
+
+/*
+  return a string representation of the calling chain for the given
+  ldb request
+ */
+char *ldb_module_call_chain(struct ldb_request *req, TALLOC_CTX *mem_ctx)
+{
+	char *ret;
+	unsigned int i = 0;
+
+	ret = talloc_strdup(mem_ctx, "");
+	if (ret == NULL) {
+		return NULL;
+	}
+
+	while (req && req->handle) {
+		char *s = talloc_asprintf_append_buffer(ret, "req[%u] %p  : %s\n",
+							i++, req, ldb_req_location(req));
+		if (s == NULL) {
+			talloc_free(ret);
+			return NULL;
+		}
+		ret = s;
+		req = req->handle->parent;
+	}
+	return ret;
+}
+
+
+/*
+  return the next module in the chain
+ */
+struct ldb_module *ldb_module_next(struct ldb_module *module)
+{
+	return module->next;
+}
+
+/*
+  set the next module in the module chain
+ */
+void ldb_module_set_next(struct ldb_module *module, struct ldb_module *next)
+{
+	module->next = next;
+}
+
+
+/*
+  get the popt_options pointer in the ldb structure. This allows a ldb
+  module to change the command line parsing
+ */
+struct poptOption **ldb_module_popt_options(struct ldb_context *ldb)
+{
+	return &ldb->popt_options;
+}
+
+
+/*
+  return the current ldb flags LDB_FLG_*
+ */
+uint32_t ldb_module_flags(struct ldb_context *ldb)
+{
+	return ldb->flags;
+}