1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-27 03:21:53 +03:00
samba-mirror/source3/winbindd/winbindd.c

1321 lines
33 KiB
C
Raw Normal View History

/*
Unix SMB/CIFS implementation.
Winbind daemon for ntdom nss module
Copyright (C) by Tim Potter 2000-2002
Copyright (C) Andrew Tridgell 2002
Copyright (C) Jelmer Vernooij 2003
Copyright (C) Volker Lendecke 2004
2008-11-26 16:01:22 +03:00
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
2008-11-26 16:01:22 +03:00
This program 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 General Public License for more details.
2008-11-26 16:01:22 +03:00
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "includes.h"
#include "winbindd.h"
#include "../../nsswitch/libwbclient/wbc_async.h"
#undef DBGC_CLASS
#define DBGC_CLASS DBGC_WINBIND
static void remove_client(struct winbindd_cli_state *state);
2009-03-12 12:12:58 +03:00
static bool opt_nocache = False;
static bool interactive = False;
extern bool override_logfile;
struct event_context *winbind_event_context(void)
{
static struct event_context *ctx;
if (!ctx && !(ctx = event_context_init(NULL))) {
smb_panic("Could not init winbind event context");
}
return ctx;
}
struct messaging_context *winbind_messaging_context(void)
{
static struct messaging_context *ctx;
if (ctx == NULL) {
ctx = messaging_init(NULL, procid_self(),
winbind_event_context());
}
if (ctx == NULL) {
DEBUG(0, ("Could not init winbind messaging context.\n"));
}
return ctx;
}
/* Reload configuration */
static bool reload_services_file(const char *lfile)
{
bool ret;
if (lp_loaded()) {
const char *fname = lp_configfile();
if (file_exist(fname) && !strcsequal(fname,get_dyn_CONFIGFILE())) {
set_dyn_CONFIGFILE(fname);
}
}
/* if this is a child, restore the logfile to the special
name - <domain>, idmap, etc. */
if (lfile && *lfile) {
lp_set_logfile(lfile);
}
reopen_logs();
ret = lp_load(get_dyn_CONFIGFILE(),False,False,True,True);
reopen_logs();
load_interfaces();
return(ret);
}
/**************************************************************************** **
Handle a fault..
**************************************************************************** */
static void fault_quit(void)
{
dump_core();
}
static void winbindd_status(void)
{
struct winbindd_cli_state *tmp;
DEBUG(0, ("winbindd status:\n"));
/* Print client state information */
2008-11-26 16:01:22 +03:00
DEBUG(0, ("\t%d clients currently active\n", winbindd_num_clients()));
2008-11-26 16:01:22 +03:00
if (DEBUGLEVEL >= 2 && winbindd_num_clients()) {
DEBUG(2, ("\tclient list:\n"));
for(tmp = winbindd_client_list(); tmp; tmp = tmp->next) {
DEBUGADD(2, ("\t\tpid %lu, sock %d\n",
(unsigned long)tmp->pid, tmp->sock));
}
}
}
/* Print winbindd status to log file */
static void print_winbindd_status(void)
{
winbindd_status();
}
/* Flush client cache */
static void flush_caches(void)
{
/* We need to invalidate cached user list entries on a SIGHUP
otherwise cached access denied errors due to restrict anonymous
hang around until the sequence number changes. */
if (!wcache_invalidate_cache()) {
DEBUG(0, ("invalidating the cache failed; revalidate the cache\n"));
if (!winbindd_cache_validate_and_initialize()) {
exit(1);
}
}
}
static void flush_caches_noinit(void)
{
/*
* We need to invalidate cached user list entries on a SIGHUP
* otherwise cached access denied errors due to restrict anonymous
* hang around until the sequence number changes.
* NB
* Skip uninitialized domains when flush cache.
* If domain is not initialized, it means it is never
* used or never become online. look, wcache_invalidate_cache()
* -> get_cache() -> init_dc_connection(). It causes a lot of traffic
* for unused domains and large traffic for primay domain's DC if there
* are many domains..
*/
if (!wcache_invalidate_cache_noinit()) {
DEBUG(0, ("invalidating the cache failed; revalidate the cache\n"));
if (!winbindd_cache_validate_and_initialize()) {
exit(1);
}
}
}
/* Handle the signal by unlinking socket and exiting */
static void terminate(bool is_parent)
{
if (is_parent) {
/* When parent goes away we should
* remove the socket file. Not so
* when children terminate.
*/
char *path = NULL;
if (asprintf(&path, "%s/%s",
get_winbind_pipe_dir(), WINBINDD_SOCKET_NAME) > 0) {
unlink(path);
SAFE_FREE(path);
}
}
idmap_close();
2008-11-26 16:01:22 +03:00
trustdom_cache_shutdown();
gencache_stabilize();
#if 0
if (interactive) {
TALLOC_CTX *mem_ctx = talloc_init("end_description");
char *description = talloc_describe_all(mem_ctx);
DEBUG(3, ("tallocs left:\n%s\n", description));
talloc_destroy(mem_ctx);
}
#endif
if (is_parent) {
s3: Fix a long-standing problem with recycled PIDs When a samba server process dies hard, it has no chance to clean up its entries in locking.tdb, brlock.tdb, connections.tdb and sessionid.tdb. For locking.tdb and brlock.tdb Samba is robust by checking every time we read an entry from the database if the corresponding process still exists. If it does not exist anymore, the entry is deleted. This is not 100% failsafe though: On systems with a limited PID space there is a non-zero chance that between the smbd's death and the fresh access, the PID is recycled by another long-running process. This renders all files that had been locked by the killed smbd potentially unusable until the new process also dies. This patch is supposed to fix the problem the following way: Every process ID in every database is augmented by a random 64-bit number that is stored in a serverid.tdb. Whenever we need to check if a process still exists we know its PID and the 64-bit number. We look up the PID in serverid.tdb and compare the 64-bit number. If it's the same, the process still is a valid smbd holding the lock. If it is different, a new smbd has taken over. I believe this is safe against an smbd that has died hard and the PID has been taken over by a non-samba process. This process would not have registered itself with a fresh 64-bit number in serverid.tdb, so the old one still exists in serverid.tdb. We protect against this case by the parent smbd taking care of deregistering PIDs from serverid.tdb and the fact that serverid.tdb is CLEAR_IF_FIRST. CLEAR_IF_FIRST does not work in a cluster, so the automatic cleanup does not work when all smbds are restarted. For this, "net serverid wipe" has to be run before smbd starts up. As a convenience, "net serverid wipedbs" also cleans up sessionid.tdb and connections.tdb. While there, this also cleans up overloading connections.tdb with all the process entries just for messaging_send_all(). Volker
2010-03-02 19:02:01 +03:00
serverid_deregister_self();
pidfile_unlink();
}
exit(0);
}
static void winbindd_sig_term_handler(struct tevent_context *ev,
struct tevent_signal *se,
int signum,
int count,
void *siginfo,
void *private_data)
{
bool *is_parent = talloc_get_type_abort(private_data, bool);
DEBUG(0,("Got sig[%d] terminate (is_parent=%d)\n",
signum, (int)*is_parent));
terminate(*is_parent);
}
bool winbindd_setup_sig_term_handler(bool parent)
{
struct tevent_signal *se;
bool *is_parent;
is_parent = talloc(winbind_event_context(), bool);
if (!is_parent) {
return false;
}
*is_parent = parent;
se = tevent_add_signal(winbind_event_context(),
is_parent,
SIGTERM, 0,
winbindd_sig_term_handler,
is_parent);
if (!se) {
DEBUG(0,("failed to setup SIGTERM handler"));
talloc_free(is_parent);
return false;
}
se = tevent_add_signal(winbind_event_context(),
is_parent,
SIGINT, 0,
winbindd_sig_term_handler,
is_parent);
if (!se) {
DEBUG(0,("failed to setup SIGINT handler"));
talloc_free(is_parent);
return false;
}
se = tevent_add_signal(winbind_event_context(),
is_parent,
SIGQUIT, 0,
winbindd_sig_term_handler,
is_parent);
if (!se) {
DEBUG(0,("failed to setup SIGINT handler"));
talloc_free(is_parent);
return false;
}
return true;
}
static void winbindd_sig_hup_handler(struct tevent_context *ev,
struct tevent_signal *se,
int signum,
int count,
void *siginfo,
void *private_data)
{
const char *file = (const char *)private_data;
DEBUG(1,("Reloading services after SIGHUP\n"));
flush_caches_noinit();
reload_services_file(file);
}
bool winbindd_setup_sig_hup_handler(const char *lfile)
{
struct tevent_signal *se;
char *file = NULL;
if (lfile) {
file = talloc_strdup(winbind_event_context(),
lfile);
if (!file) {
return false;
}
}
se = tevent_add_signal(winbind_event_context(),
winbind_event_context(),
SIGHUP, 0,
winbindd_sig_hup_handler,
file);
if (!se) {
return false;
}
return true;
}
static void winbindd_sig_chld_handler(struct tevent_context *ev,
struct tevent_signal *se,
int signum,
int count,
void *siginfo,
void *private_data)
{
pid_t pid;
while ((pid = sys_waitpid(-1, NULL, WNOHANG)) > 0) {
winbind_child_died(pid);
}
}
static bool winbindd_setup_sig_chld_handler(void)
{
struct tevent_signal *se;
se = tevent_add_signal(winbind_event_context(),
winbind_event_context(),
SIGCHLD, 0,
winbindd_sig_chld_handler,
NULL);
if (!se) {
return false;
}
return true;
}
static void winbindd_sig_usr2_handler(struct tevent_context *ev,
struct tevent_signal *se,
int signum,
int count,
void *siginfo,
void *private_data)
{
print_winbindd_status();
}
static bool winbindd_setup_sig_usr2_handler(void)
{
struct tevent_signal *se;
se = tevent_add_signal(winbind_event_context(),
winbind_event_context(),
SIGUSR2, 0,
winbindd_sig_usr2_handler,
NULL);
if (!se) {
return false;
}
return true;
}
/* React on 'smbcontrol winbindd reload-config' in the same way as on SIGHUP*/
static void msg_reload_services(struct messaging_context *msg,
void *private_data,
uint32_t msg_type,
struct server_id server_id,
DATA_BLOB *data)
{
/* Flush various caches */
flush_caches();
reload_services_file((const char *) private_data);
}
/* React on 'smbcontrol winbindd shutdown' in the same way as on SIGTERM*/
static void msg_shutdown(struct messaging_context *msg,
void *private_data,
uint32_t msg_type,
struct server_id server_id,
DATA_BLOB *data)
{
/* only the parent waits for this message */
DEBUG(0,("Got shutdown message\n"));
terminate(true);
}
static void winbind_msg_validate_cache(struct messaging_context *msg_ctx,
void *private_data,
uint32_t msg_type,
struct server_id server_id,
DATA_BLOB *data)
{
uint8 ret;
pid_t child_pid;
struct sigaction act;
struct sigaction oldact;
DEBUG(10, ("winbindd_msg_validate_cache: got validate-cache "
"message.\n"));
/*
* call the validation code from a child:
* so we don't block the main winbindd and the validation
* code can safely use fork/waitpid...
*/
CatchChild();
child_pid = sys_fork();
if (child_pid == -1) {
DEBUG(1, ("winbind_msg_validate_cache: Could not fork: %s\n",
strerror(errno)));
return;
}
if (child_pid != 0) {
/* parent */
DEBUG(5, ("winbind_msg_validate_cache: child created with "
"pid %d.\n", (int)child_pid));
return;
}
/* child */
/* install default SIGCHLD handler: validation code uses fork/waitpid */
ZERO_STRUCT(act);
act.sa_handler = SIG_DFL;
#ifdef SA_RESTART
/* We *want* SIGALRM to interrupt a system call. */
act.sa_flags = SA_RESTART;
#endif
sigemptyset(&act.sa_mask);
sigaddset(&act.sa_mask,SIGCHLD);
sigaction(SIGCHLD,&act,&oldact);
ret = (uint8)winbindd_validate_cache_nobackup();
DEBUG(10, ("winbindd_msg_validata_cache: got return value %d\n", ret));
messaging_send_buf(msg_ctx, server_id, MSG_WINBIND_VALIDATE_CACHE, &ret,
(size_t)1);
_exit(0);
}
static struct winbindd_dispatch_table {
enum winbindd_cmd cmd;
void (*fn)(struct winbindd_cli_state *state);
const char *winbindd_cmd_name;
} dispatch_table[] = {
2008-11-26 16:01:22 +03:00
/* PAM auth functions */
{ WINBINDD_PAM_AUTH, winbindd_pam_auth, "PAM_AUTH" },
{ WINBINDD_PAM_AUTH_CRAP, winbindd_pam_auth_crap, "AUTH_CRAP" },
{ WINBINDD_PAM_CHAUTHTOK, winbindd_pam_chauthtok, "CHAUTHTOK" },
{ WINBINDD_PAM_LOGOFF, winbindd_pam_logoff, "PAM_LOGOFF" },
{ WINBINDD_PAM_CHNG_PSWD_AUTH_CRAP, winbindd_pam_chng_pswd_auth_crap, "CHNG_PSWD_AUTH_CRAP" },
/* Enumeration functions */
{ WINBINDD_LIST_TRUSTDOM, winbindd_list_trusted_domains,
"LIST_TRUSTDOM" },
/* Miscellaneous */
{ WINBINDD_INFO, winbindd_info, "INFO" },
{ WINBINDD_INTERFACE_VERSION, winbindd_interface_version,
"INTERFACE_VERSION" },
{ WINBINDD_DOMAIN_NAME, winbindd_domain_name, "DOMAIN_NAME" },
{ WINBINDD_DOMAIN_INFO, winbindd_domain_info, "DOMAIN_INFO" },
{ WINBINDD_NETBIOS_NAME, winbindd_netbios_name, "NETBIOS_NAME" },
{ WINBINDD_PRIV_PIPE_DIR, winbindd_priv_pipe_dir,
"WINBINDD_PRIV_PIPE_DIR" },
/* Credential cache access */
{ WINBINDD_CCACHE_NTLMAUTH, winbindd_ccache_ntlm_auth, "NTLMAUTH" },
{ WINBINDD_CCACHE_SAVE, winbindd_ccache_save, "CCACHE_SAVE" },
/* WINS functions */
{ WINBINDD_WINS_BYNAME, winbindd_wins_byname, "WINS_BYNAME" },
{ WINBINDD_WINS_BYIP, winbindd_wins_byip, "WINS_BYIP" },
/* End of list */
{ WINBINDD_NUM_CMDS, NULL, "NONE" }
};
2009-05-16 14:00:07 +04:00
struct winbindd_async_dispatch_table {
enum winbindd_cmd cmd;
const char *cmd_name;
struct tevent_req *(*send_req)(TALLOC_CTX *mem_ctx,
struct tevent_context *ev,
struct winbindd_cli_state *cli,
2009-05-16 14:00:07 +04:00
struct winbindd_request *request);
NTSTATUS (*recv_req)(struct tevent_req *req,
struct winbindd_response *presp);
2009-05-16 14:00:07 +04:00
};
static struct winbindd_async_dispatch_table async_nonpriv_table[] = {
2009-05-16 14:00:34 +04:00
{ WINBINDD_PING, "PING",
wb_ping_send, wb_ping_recv },
{ WINBINDD_LOOKUPSID, "LOOKUPSID",
winbindd_lookupsid_send, winbindd_lookupsid_recv },
{ WINBINDD_LOOKUPNAME, "LOOKUPNAME",
winbindd_lookupname_send, winbindd_lookupname_recv },
{ WINBINDD_SID_TO_UID, "SID_TO_UID",
winbindd_sid_to_uid_send, winbindd_sid_to_uid_recv },
{ WINBINDD_SID_TO_GID, "SID_TO_GID",
winbindd_sid_to_gid_send, winbindd_sid_to_gid_recv },
{ WINBINDD_UID_TO_SID, "UID_TO_SID",
winbindd_uid_to_sid_send, winbindd_uid_to_sid_recv },
{ WINBINDD_GID_TO_SID, "GID_TO_SID",
winbindd_gid_to_sid_send, winbindd_gid_to_sid_recv },
{ WINBINDD_GETPWSID, "GETPWSID",
winbindd_getpwsid_send, winbindd_getpwsid_recv },
{ WINBINDD_GETPWNAM, "GETPWNAM",
winbindd_getpwnam_send, winbindd_getpwnam_recv },
{ WINBINDD_GETPWUID, "GETPWUID",
winbindd_getpwuid_send, winbindd_getpwuid_recv },
{ WINBINDD_GETSIDALIASES, "GETSIDALIASES",
winbindd_getsidaliases_send, winbindd_getsidaliases_recv },
{ WINBINDD_GETUSERDOMGROUPS, "GETUSERDOMGROUPS",
winbindd_getuserdomgroups_send, winbindd_getuserdomgroups_recv },
{ WINBINDD_GETGROUPS, "GETGROUPS",
winbindd_getgroups_send, winbindd_getgroups_recv },
{ WINBINDD_SHOW_SEQUENCE, "SHOW_SEQUENCE",
winbindd_show_sequence_send, winbindd_show_sequence_recv },
{ WINBINDD_GETGRGID, "GETGRGID",
winbindd_getgrgid_send, winbindd_getgrgid_recv },
{ WINBINDD_GETGRNAM, "GETGRNAM",
winbindd_getgrnam_send, winbindd_getgrnam_recv },
{ WINBINDD_GETUSERSIDS, "GETUSERSIDS",
winbindd_getusersids_send, winbindd_getusersids_recv },
{ WINBINDD_LOOKUPRIDS, "LOOKUPRIDS",
winbindd_lookuprids_send, winbindd_lookuprids_recv },
{ WINBINDD_SETPWENT, "SETPWENT",
winbindd_setpwent_send, winbindd_setpwent_recv },
{ WINBINDD_GETPWENT, "GETPWENT",
winbindd_getpwent_send, winbindd_getpwent_recv },
{ WINBINDD_ENDPWENT, "ENDPWENT",
winbindd_endpwent_send, winbindd_endpwent_recv },
{ WINBINDD_DSGETDCNAME, "DSGETDCNAME",
winbindd_dsgetdcname_send, winbindd_dsgetdcname_recv },
{ WINBINDD_GETDCNAME, "GETDCNAME",
winbindd_getdcname_send, winbindd_getdcname_recv },
{ WINBINDD_SETGRENT, "SETGRENT",
winbindd_setgrent_send, winbindd_setgrent_recv },
{ WINBINDD_GETGRENT, "GETGRENT",
winbindd_getgrent_send, winbindd_getgrent_recv },
{ WINBINDD_ENDGRENT, "ENDGRENT",
winbindd_endgrent_send, winbindd_endgrent_recv },
{ WINBINDD_LIST_USERS, "LIST_USERS",
winbindd_list_users_send, winbindd_list_users_recv },
{ WINBINDD_LIST_GROUPS, "LIST_GROUPS",
winbindd_list_groups_send, winbindd_list_groups_recv },
{ WINBINDD_CHECK_MACHACC, "CHECK_MACHACC",
winbindd_check_machine_acct_send, winbindd_check_machine_acct_recv },
{ WINBINDD_PING_DC, "PING_DC",
winbindd_ping_dc_send, winbindd_ping_dc_recv },
2009-05-16 14:00:34 +04:00
2009-05-16 14:00:07 +04:00
{ 0, NULL, NULL, NULL }
};
static struct winbindd_async_dispatch_table async_priv_table[] = {
{ WINBINDD_ALLOCATE_UID, "ALLOCATE_UID",
winbindd_allocate_uid_send, winbindd_allocate_uid_recv },
{ WINBINDD_ALLOCATE_GID, "ALLOCATE_GID",
winbindd_allocate_gid_send, winbindd_allocate_gid_recv },
{ WINBINDD_SET_MAPPING, "SET_MAPPING",
winbindd_set_mapping_send, winbindd_set_mapping_recv },
{ WINBINDD_REMOVE_MAPPING, "SET_MAPPING",
winbindd_remove_mapping_send, winbindd_remove_mapping_recv },
{ WINBINDD_SET_HWM, "SET_HWM",
winbindd_set_hwm_send, winbindd_set_hwm_recv },
{ WINBINDD_CHANGE_MACHACC, "CHANGE_MACHACC",
winbindd_change_machine_acct_send, winbindd_change_machine_acct_recv },
{ 0, NULL, NULL, NULL }
};
2009-05-16 14:00:07 +04:00
static void wb_request_done(struct tevent_req *req);
static void process_request(struct winbindd_cli_state *state)
{
struct winbindd_dispatch_table *table = dispatch_table;
2009-05-16 14:00:07 +04:00
struct winbindd_async_dispatch_table *atable;
state->mem_ctx = talloc_init("winbind request");
if (state->mem_ctx == NULL)
return;
/* Remember who asked us. */
state->pid = state->request->pid;
state->cmd_name = "unknown request";
state->recv_fn = NULL;
/* Process command */
2009-05-16 14:00:07 +04:00
for (atable = async_nonpriv_table; atable->send_req; atable += 1) {
if (state->request->cmd == atable->cmd) {
break;
}
}
if ((atable->send_req == NULL) && state->privileged) {
for (atable = async_priv_table; atable->send_req;
atable += 1) {
if (state->request->cmd == atable->cmd) {
break;
}
}
}
2009-05-16 14:00:07 +04:00
if (atable->send_req != NULL) {
struct tevent_req *req;
state->cmd_name = atable->cmd_name;
state->recv_fn = atable->recv_req;
DEBUG(10, ("process_request: Handling async request %d:%s\n",
(int)state->pid, state->cmd_name));
2009-05-16 14:00:07 +04:00
req = atable->send_req(state->mem_ctx, winbind_event_context(),
state, state->request);
2009-05-16 14:00:07 +04:00
if (req == NULL) {
DEBUG(0, ("process_request: atable->send failed for "
"%s\n", atable->cmd_name));
request_error(state);
return;
}
tevent_req_set_callback(req, wb_request_done, state);
return;
}
state->response = talloc_zero(state->mem_ctx,
struct winbindd_response);
if (state->response == NULL) {
DEBUG(10, ("talloc failed\n"));
remove_client(state);
return;
}
state->response->result = WINBINDD_PENDING;
state->response->length = sizeof(struct winbindd_response);
for (table = dispatch_table; table->fn; table++) {
if (state->request->cmd == table->cmd) {
DEBUG(10,("process_request: request fn %s\n",
table->winbindd_cmd_name ));
state->cmd_name = table->winbindd_cmd_name;
table->fn(state);
break;
}
}
if (!table->fn) {
DEBUG(10,("process_request: unknown request fn number %d\n",
(int)state->request->cmd ));
request_error(state);
}
}
2009-05-16 14:00:07 +04:00
static void wb_request_done(struct tevent_req *req)
{
struct winbindd_cli_state *state = tevent_req_callback_data(
req, struct winbindd_cli_state);
NTSTATUS status;
2010-03-22 17:29:19 +03:00
state->response = talloc_zero(state->mem_ctx,
struct winbindd_response);
if (state->response == NULL) {
DEBUG(0, ("wb_request_done[%d:%s]: talloc_zero failed - removing client\n",
(int)state->pid, state->cmd_name));
remove_client(state);
return;
}
state->response->result = WINBINDD_PENDING;
state->response->length = sizeof(struct winbindd_response);
status = state->recv_fn(req, state->response);
2009-05-16 14:00:07 +04:00
TALLOC_FREE(req);
DEBUG(10,("wb_request_done[%d:%s]: %s\n",
(int)state->pid, state->cmd_name, nt_errstr(status)));
2009-05-16 14:00:07 +04:00
if (!NT_STATUS_IS_OK(status)) {
request_error(state);
return;
2009-05-16 14:00:07 +04:00
}
request_ok(state);
}
/*
* This is the main event loop of winbind requests. It goes through a
* state-machine of 3 read/write requests, 4 if you have extra data to send.
*
* An idle winbind client has a read request of 4 bytes outstanding,
* finalizing function is request_len_recv, checking the length. request_recv
* then processes the packet. The processing function then at some point has
* to call request_finished which schedules sending the response.
*/
static void request_finished(struct winbindd_cli_state *state);
static void winbind_client_request_read(struct tevent_req *req);
static void winbind_client_response_written(struct tevent_req *req);
static void request_finished(struct winbindd_cli_state *state)
{
struct tevent_req *req;
TALLOC_FREE(state->request);
req = wb_resp_write_send(state, winbind_event_context(),
state->out_queue, state->sock,
state->response);
if (req == NULL) {
DEBUG(10,("request_finished[%d:%s]: wb_resp_write_send() failed\n",
(int)state->pid, state->cmd_name));
remove_client(state);
return;
}
tevent_req_set_callback(req, winbind_client_response_written, state);
}
static void winbind_client_response_written(struct tevent_req *req)
{
struct winbindd_cli_state *state = tevent_req_callback_data(
req, struct winbindd_cli_state);
ssize_t ret;
int err;
ret = wb_resp_write_recv(req, &err);
TALLOC_FREE(req);
if (ret == -1) {
close(state->sock);
state->sock = -1;
DEBUG(2, ("Could not write response[%d:%s] to client: %s\n",
(int)state->pid, state->cmd_name, strerror(err)));
remove_client(state);
return;
}
DEBUG(10,("winbind_client_response_written[%d:%s]: deliverd response to client\n",
(int)state->pid, state->cmd_name));
TALLOC_FREE(state->mem_ctx);
state->response = NULL;
state->cmd_name = "no request";
state->recv_fn = NULL;
req = wb_req_read_send(state, winbind_event_context(), state->sock,
WINBINDD_MAX_EXTRA_DATA);
if (req == NULL) {
remove_client(state);
return;
}
tevent_req_set_callback(req, winbind_client_request_read, state);
}
void request_error(struct winbindd_cli_state *state)
{
SMB_ASSERT(state->response->result == WINBINDD_PENDING);
state->response->result = WINBINDD_ERROR;
request_finished(state);
}
void request_ok(struct winbindd_cli_state *state)
{
SMB_ASSERT(state->response->result == WINBINDD_PENDING);
state->response->result = WINBINDD_OK;
request_finished(state);
}
/* Process a new connection by adding it to the client connection list */
static void new_connection(int listen_sock, bool privileged)
{
struct sockaddr_un sunaddr;
struct winbindd_cli_state *state;
struct tevent_req *req;
socklen_t len;
int sock;
2008-11-26 16:01:22 +03:00
/* Accept connection */
2008-11-26 16:01:22 +03:00
len = sizeof(sunaddr);
do {
2009-05-08 01:07:55 +04:00
sock = accept(listen_sock, (struct sockaddr *)(void *)&sunaddr,
&len);
} while (sock == -1 && errno == EINTR);
if (sock == -1)
return;
2008-11-26 16:01:22 +03:00
DEBUG(6,("accepted socket %d\n", sock));
2008-11-26 16:01:22 +03:00
/* Create new connection structure */
2008-11-26 16:01:22 +03:00
if ((state = TALLOC_ZERO_P(NULL, struct winbindd_cli_state)) == NULL) {
close(sock);
return;
}
2008-11-26 16:01:22 +03:00
state->sock = sock;
state->out_queue = tevent_queue_create(state, "winbind client reply");
if (state->out_queue == NULL) {
close(sock);
TALLOC_FREE(state);
return;
}
state->last_access = time(NULL);
state->privileged = privileged;
2003-03-24 12:54:13 +03:00
req = wb_req_read_send(state, winbind_event_context(), state->sock,
WINBINDD_MAX_EXTRA_DATA);
if (req == NULL) {
TALLOC_FREE(state);
close(sock);
return;
}
tevent_req_set_callback(req, winbind_client_request_read, state);
/* Add to connection list */
2008-11-26 16:01:22 +03:00
winbindd_add_client(state);
}
static void winbind_client_request_read(struct tevent_req *req)
{
struct winbindd_cli_state *state = tevent_req_callback_data(
req, struct winbindd_cli_state);
ssize_t ret;
int err;
ret = wb_req_read_recv(req, state, &state->request, &err);
2009-07-25 03:03:45 +04:00
TALLOC_FREE(req);
if (ret == -1) {
if (err == EPIPE) {
DEBUG(6, ("closing socket %d, client exited\n",
state->sock));
} else {
DEBUG(2, ("Could not read client request from fd %d: "
"%s\n", state->sock, strerror(err)));
}
close(state->sock);
state->sock = -1;
remove_client(state);
return;
}
process_request(state);
}
/* Remove a client connection from client connection list */
static void remove_client(struct winbindd_cli_state *state)
{
char c = 0;
int nwritten;
/* It's a dead client - hold a funeral */
2008-11-26 16:01:22 +03:00
if (state == NULL) {
return;
}
if (state->sock != -1) {
/* tell client, we are closing ... */
nwritten = write(state->sock, &c, sizeof(c));
if (nwritten == -1) {
DEBUG(2, ("final write to client failed: %s\n",
strerror(errno)));
}
/* Close socket */
2008-11-26 16:01:22 +03:00
close(state->sock);
state->sock = -1;
}
2008-11-26 16:01:22 +03:00
TALLOC_FREE(state->mem_ctx);
/* Remove from list and free */
2008-11-26 16:01:22 +03:00
winbindd_remove_client(state);
TALLOC_FREE(state);
}
/* Shutdown client connection which has been idle for the longest time */
static bool remove_idle_client(void)
{
struct winbindd_cli_state *state, *remove_state = NULL;
time_t last_access = 0;
int nidle = 0;
for (state = winbindd_client_list(); state; state = state->next) {
if (state->response == NULL &&
!state->pwent_state && !state->grent_state) {
nidle++;
if (!last_access || state->last_access < last_access) {
last_access = state->last_access;
remove_state = state;
}
}
}
if (remove_state) {
DEBUG(5,("Found %d idle client connections, shutting down sock %d, pid %u\n",
nidle, remove_state->sock, (unsigned int)remove_state->pid));
remove_client(remove_state);
return True;
}
return False;
}
struct winbindd_listen_state {
bool privileged;
int fd;
};
static void winbindd_listen_fde_handler(struct tevent_context *ev,
struct tevent_fd *fde,
uint16_t flags,
void *private_data)
{
struct winbindd_listen_state *s = talloc_get_type_abort(private_data,
struct winbindd_listen_state);
while (winbindd_num_clients() >
WINBINDD_MAX_SIMULTANEOUS_CLIENTS - 1) {
DEBUG(5,("winbindd: Exceeding %d client "
"connections, removing idle "
"connection.\n",
WINBINDD_MAX_SIMULTANEOUS_CLIENTS));
if (!remove_idle_client()) {
DEBUG(0,("winbindd: Exceeding %d "
"client connections, no idle "
"connection found\n",
WINBINDD_MAX_SIMULTANEOUS_CLIENTS));
break;
}
}
new_connection(s->fd, s->privileged);
}
static bool winbindd_setup_listeners(void)
{
struct winbindd_listen_state *pub_state = NULL;
struct winbindd_listen_state *priv_state = NULL;
2009-05-25 22:31:59 +04:00
struct tevent_fd *fde;
pub_state = talloc(winbind_event_context(),
struct winbindd_listen_state);
if (!pub_state) {
goto failed;
}
pub_state->privileged = false;
pub_state->fd = open_winbindd_socket();
if (pub_state->fd == -1) {
goto failed;
}
2009-05-25 22:31:59 +04:00
fde = tevent_add_fd(winbind_event_context(), pub_state, pub_state->fd,
TEVENT_FD_READ, winbindd_listen_fde_handler,
pub_state);
if (fde == NULL) {
close(pub_state->fd);
goto failed;
}
2009-05-25 22:31:59 +04:00
tevent_fd_set_auto_close(fde);
priv_state = talloc(winbind_event_context(),
struct winbindd_listen_state);
if (!priv_state) {
goto failed;
}
priv_state->privileged = true;
priv_state->fd = open_winbindd_priv_socket();
if (priv_state->fd == -1) {
goto failed;
}
2009-05-25 22:31:59 +04:00
fde = tevent_add_fd(winbind_event_context(), priv_state,
priv_state->fd, TEVENT_FD_READ,
winbindd_listen_fde_handler, priv_state);
if (fde == NULL) {
close(priv_state->fd);
goto failed;
}
2009-05-25 22:31:59 +04:00
tevent_fd_set_auto_close(fde);
return true;
failed:
TALLOC_FREE(pub_state);
TALLOC_FREE(priv_state);
return false;
}
2009-03-12 12:12:58 +03:00
bool winbindd_use_idmap_cache(void)
{
return !opt_nocache;
}
bool winbindd_use_cache(void)
{
return !opt_nocache;
}
/* Main function */
int main(int argc, char **argv, char **envp)
{
static bool is_daemon = False;
static bool Fork = True;
static bool log_stdout = False;
static bool no_process_group = False;
enum {
OPT_DAEMON = 1000,
OPT_FORK,
OPT_NO_PROCESS_GROUP,
OPT_LOG_STDOUT
};
struct poptOption long_options[] = {
POPT_AUTOHELP
{ "stdout", 'S', POPT_ARG_NONE, NULL, OPT_LOG_STDOUT, "Log to stdout" },
{ "foreground", 'F', POPT_ARG_NONE, NULL, OPT_FORK, "Daemon in foreground mode" },
{ "no-process-group", 0, POPT_ARG_NONE, NULL, OPT_NO_PROCESS_GROUP, "Don't create a new process group" },
{ "daemon", 'D', POPT_ARG_NONE, NULL, OPT_DAEMON, "Become a daemon (default)" },
{ "interactive", 'i', POPT_ARG_NONE, NULL, 'i', "Interactive mode" },
{ "no-caching", 'n', POPT_ARG_NONE, NULL, 'n', "Disable caching" },
POPT_COMMON_SAMBA
POPT_TABLEEND
};
poptContext pc;
int opt;
TALLOC_CTX *frame = talloc_stackframe();
struct tevent_timer *te;
/* glibc (?) likes to print "User defined signal 1" and exit if a
SIGUSR[12] is received before a handler is installed */
CatchSignal(SIGUSR1, SIG_IGN);
CatchSignal(SIGUSR2, SIG_IGN);
fault_setup((void (*)(void *))fault_quit );
dump_core_setup("winbindd");
load_case_tables();
/* Initialise for running in non-root mode */
sec_init();
set_remote_machine_name("winbindd", False);
/* Set environment variable so we don't recursively call ourselves.
This may also be useful interactively. */
if ( !winbind_off() ) {
DEBUG(0,("Failed to disable recusive winbindd calls. Exiting.\n"));
exit(1);
}
/* Initialise samba/rpc client stuff */
pc = poptGetContext("winbindd", argc, (const char **)argv, long_options, 0);
while ((opt = poptGetNextOpt(pc)) != -1) {
switch (opt) {
/* Don't become a daemon */
case OPT_DAEMON:
is_daemon = True;
break;
case 'i':
interactive = True;
log_stdout = True;
Fork = False;
break;
case OPT_FORK:
Fork = false;
break;
case OPT_NO_PROCESS_GROUP:
no_process_group = true;
break;
case OPT_LOG_STDOUT:
log_stdout = true;
break;
case 'n':
opt_nocache = true;
break;
default:
d_fprintf(stderr, "\nInvalid option %s: %s\n\n",
poptBadOption(pc, 0), poptStrerror(opt));
poptPrintUsage(pc, stderr, 0);
exit(1);
}
}
if (is_daemon && interactive) {
d_fprintf(stderr,"\nERROR: "
"Option -i|--interactive is not allowed together with -D|--daemon\n\n");
poptPrintUsage(pc, stderr, 0);
exit(1);
}
if (log_stdout && Fork) {
d_fprintf(stderr, "\nERROR: "
"Can't log to stdout (-S) unless daemon is in foreground +(-F) or interactive (-i)\n\n");
poptPrintUsage(pc, stderr, 0);
exit(1);
}
poptFreeContext(pc);
if (!override_logfile) {
char *lfile = NULL;
if (asprintf(&lfile,"%s/log.winbindd",
get_dyn_LOGFILEBASE()) > 0) {
lp_set_logfile(lfile);
SAFE_FREE(lfile);
}
}
setup_logging("winbindd", log_stdout);
reopen_logs();
DEBUG(0,("winbindd version %s started.\n", samba_version_string()));
DEBUGADD(0,("%s\n", COPYRIGHT_STARTUP_MESSAGE));
if (!lp_load_initial_only(get_dyn_CONFIGFILE())) {
DEBUG(0, ("error opening config file\n"));
exit(1);
}
/* Initialise messaging system */
if (winbind_messaging_context() == NULL) {
exit(1);
}
if (!reload_services_file(NULL)) {
DEBUG(0, ("error opening config file\n"));
exit(1);
}
if (!directory_exist(lp_lockdir())) {
mkdir(lp_lockdir(), 0755);
}
/* Setup names. */
if (!init_names())
exit(1);
load_interfaces();
if (!secrets_init()) {
DEBUG(0,("Could not initialize domain trust account secrets. Giving up\n"));
return False;
}
/* Enable netbios namecache */
namecache_enable();
/* Unblock all signals we are interested in as they may have been
blocked by the parent process. */
BlockSignals(False, SIGINT);
BlockSignals(False, SIGQUIT);
BlockSignals(False, SIGTERM);
BlockSignals(False, SIGUSR1);
BlockSignals(False, SIGUSR2);
BlockSignals(False, SIGHUP);
BlockSignals(False, SIGCHLD);
if (!interactive)
become_daemon(Fork, no_process_group, log_stdout);
pidfile_create("winbindd");
#if HAVE_SETPGID
/*
* If we're interactive we want to set our own process group for
* signal management.
*/
if (interactive && !no_process_group)
setpgid( (pid_t)0, (pid_t)0);
#endif
TimeInit();
/* Don't use winbindd_reinit_after_fork here as
* we're just starting up and haven't created any
* winbindd-specific resources we must free yet. JRA.
*/
if (!NT_STATUS_IS_OK(reinit_after_fork(winbind_messaging_context(),
winbind_event_context(),
false))) {
DEBUG(0,("reinit_after_fork() failed\n"));
exit(1);
}
/* Setup signal handlers */
if (!winbindd_setup_sig_term_handler(true))
exit(1);
if (!winbindd_setup_sig_hup_handler(NULL))
exit(1);
if (!winbindd_setup_sig_chld_handler())
exit(1);
if (!winbindd_setup_sig_usr2_handler())
exit(1);
CatchSignal(SIGPIPE, SIG_IGN); /* Ignore sigpipe */
/*
* Ensure all cache and idmap caches are consistent
* and initialized before we startup.
*/
if (!winbindd_cache_validate_and_initialize()) {
exit(1);
}
/* get broadcast messages */
s3: Fix a long-standing problem with recycled PIDs When a samba server process dies hard, it has no chance to clean up its entries in locking.tdb, brlock.tdb, connections.tdb and sessionid.tdb. For locking.tdb and brlock.tdb Samba is robust by checking every time we read an entry from the database if the corresponding process still exists. If it does not exist anymore, the entry is deleted. This is not 100% failsafe though: On systems with a limited PID space there is a non-zero chance that between the smbd's death and the fresh access, the PID is recycled by another long-running process. This renders all files that had been locked by the killed smbd potentially unusable until the new process also dies. This patch is supposed to fix the problem the following way: Every process ID in every database is augmented by a random 64-bit number that is stored in a serverid.tdb. Whenever we need to check if a process still exists we know its PID and the 64-bit number. We look up the PID in serverid.tdb and compare the 64-bit number. If it's the same, the process still is a valid smbd holding the lock. If it is different, a new smbd has taken over. I believe this is safe against an smbd that has died hard and the PID has been taken over by a non-samba process. This process would not have registered itself with a fresh 64-bit number in serverid.tdb, so the old one still exists in serverid.tdb. We protect against this case by the parent smbd taking care of deregistering PIDs from serverid.tdb and the fact that serverid.tdb is CLEAR_IF_FIRST. CLEAR_IF_FIRST does not work in a cluster, so the automatic cleanup does not work when all smbds are restarted. For this, "net serverid wipe" has to be run before smbd starts up. As a convenience, "net serverid wipedbs" also cleans up sessionid.tdb and connections.tdb. While there, this also cleans up overloading connections.tdb with all the process entries just for messaging_send_all(). Volker
2010-03-02 19:02:01 +03:00
if (!serverid_register_self(FLAG_MSG_GENERAL|FLAG_MSG_DBWRAP)) {
DEBUG(1, ("Could not register myself in serverid.tdb\n"));
exit(1);
}
/* React on 'smbcontrol winbindd reload-config' in the same way
as to SIGHUP signal */
messaging_register(winbind_messaging_context(), NULL,
MSG_SMB_CONF_UPDATED, msg_reload_services);
messaging_register(winbind_messaging_context(), NULL,
MSG_SHUTDOWN, msg_shutdown);
/* Handle online/offline messages. */
messaging_register(winbind_messaging_context(), NULL,
MSG_WINBIND_OFFLINE, winbind_msg_offline);
messaging_register(winbind_messaging_context(), NULL,
MSG_WINBIND_ONLINE, winbind_msg_online);
messaging_register(winbind_messaging_context(), NULL,
MSG_WINBIND_ONLINESTATUS, winbind_msg_onlinestatus);
messaging_register(winbind_messaging_context(), NULL,
MSG_DUMP_EVENT_LIST, winbind_msg_dump_event_list);
messaging_register(winbind_messaging_context(), NULL,
MSG_WINBIND_VALIDATE_CACHE,
winbind_msg_validate_cache);
messaging_register(winbind_messaging_context(), NULL,
MSG_WINBIND_DUMP_DOMAIN_LIST,
winbind_msg_dump_domain_list);
/* Register handler for MSG_DEBUG. */
messaging_register(winbind_messaging_context(), NULL,
MSG_DEBUG,
winbind_msg_debug);
2008-11-26 16:01:22 +03:00
netsamlogon_cache_init(); /* Non-critical */
2008-11-26 16:01:22 +03:00
/* clear the cached list of trusted domains */
wcache_tdc_clear();
2008-11-26 16:01:22 +03:00
if (!init_domain_list()) {
DEBUG(0,("unable to initialize domain list\n"));
exit(1);
}
init_idmap_child();
init_locator_child();
smb_nscd_flush_user_cache();
smb_nscd_flush_group_cache();
/* setup listen sockets */
if (!winbindd_setup_listeners()) {
DEBUG(0,("winbindd_setup_listeners() failed\n"));
exit(1);
}
te = tevent_add_timer(winbind_event_context(), NULL, timeval_zero(),
rescan_trusted_domains, NULL);
if (te == NULL) {
DEBUG(0, ("Could not trigger rescan_trusted_domains()\n"));
exit(1);
}
TALLOC_FREE(frame);
/* Loop waiting for requests */
while (1) {
frame = talloc_stackframe();
if (tevent_loop_once(winbind_event_context()) == -1) {
DEBUG(1, ("tevent_loop_once() failed: %s\n",
strerror(errno)));
return 1;
}
TALLOC_FREE(frame);
}
return 0;
}