sin/samba-mirror
1
0
Fork 0
mirror of https://github.com/samba-team/samba.git synced 2025-01-15 23:24:37 +03:00
Code
samba-mirror/source3/nsswitch/winbindd_acct.c

1206 lines
30 KiB
C
Raw Normal View History

Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
/*
Unix SMB/CIFS implementation.
Winbind account management functions
Copyright (C) by Gerald (Jerry) Carter 2003
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 2 of the License, or
(at your option) any later version.
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.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
#include "winbindd.h"
#undef DBGC_CLASS
#define DBGC_CLASS DBGC_WINBIND
#define WBKEY_PASSWD "WBA_PASSWD"
#define WBKEY_GROUP "WBA_GROUP"
#define NUM_PW_FIELDS 7
#define NUM_GRP_FIELDS 4
/* Globals */
static TDB_CONTEXT *account_tdb;
extern userdom_struct current_user_info;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
struct _check_primary_grp {
gid_t gid;
BOOL found;
};
/**********************************************************************
**********************************************************************/
static void free_winbindd_gr( WINBINDD_GR *grp )
{
int i;
if ( !grp )
return;
for ( i=0; i<grp->num_gr_mem; i++ )
SAFE_FREE( grp->gr_mem[i] );
SAFE_FREE( grp->gr_mem );
return;
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
/*****************************************************************************
Initialise auto-account database.
*****************************************************************************/
static BOOL winbindd_accountdb_init(void)
{
/* see if we've already opened the tdb */
if ( account_tdb )
return True;
rework winbindd_accountdb_init() to reduce error messages in the log (This used to be commit d98a68e0ebaf2fbd360b826f5df472bc3f94285b)
2003-08-06 18:01:39 +00:00
/* winbindd_idmap.tdb should always be opened by the idmap_init()
code first */
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
rework winbindd_accountdb_init() to reduce error messages in the log (This used to be commit d98a68e0ebaf2fbd360b826f5df472bc3f94285b)
2003-08-06 18:01:39 +00:00
if ( !(account_tdb = idmap_tdb_handle()) ) {
DEBUG(0, ("winbindd_accountdb_init: Unable to retreive handle for database\n"));
return False;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
/* yeah! */
return True;
}
/**********************************************************************
Convert a string in /etc/passwd format to a struct passwd* entry
**********************************************************************/
static WINBINDD_PW* string2passwd( char *string )
{
static WINBINDD_PW pw;
char *p, *str;
char *fields[NUM_PW_FIELDS];
int i;
if ( !string )
return NULL;
ZERO_STRUCTP( &pw );
DEBUG(10,("string2passwd: converting \"%s\"\n", string));
ZERO_STRUCT( fields );
for ( i=0, str=string; i<NUM_PW_FIELDS-1; i++ ) {
if ( !(p = strchr( str, ':' )) ) {
DEBUG(0,("string2passwd: parsing failure\n"));
return NULL;
}
*p = '\0';
if ( str )
fields[i] = str;
str = p + 1;
}
if ( str )
fields[i] = str;
/* copy fields */
fstrcpy( pw.pw_name, fields[0] );
fstrcpy( pw.pw_passwd, fields[1] );
pw.pw_uid = atoi( fields[2] );
pw.pw_gid = atoi( fields[3] );
fstrcpy( pw.pw_gecos, fields[4] );
fstrcpy( pw.pw_dir, fields[5] );
fstrcpy( pw.pw_shell, fields[6] );
/* last minute sanity checks */
if ( pw.pw_uid==0 || pw.pw_gid==0 ) {
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(0,("string2passwd: Failure! uid==%lu, gid==%lu\n",
(unsigned long)pw.pw_uid, (unsigned long)pw.pw_gid));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return NULL;
}
DEBUG(10,("string2passwd: Success\n"));
return &pw;
}
/**********************************************************************
Convert a struct passwd* to a string formatted for /etc/passwd
**********************************************************************/
static char* passwd2string( const WINBINDD_PW *pw )
{
static pstring string;
int ret;
if ( !pw || !pw->pw_name )
return NULL;
DEBUG(10,("passwd2string: converting passwd struct for %s\n",
pw->pw_name));
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
ret = pstr_sprintf( string, "%s:%s:%lu:%lu:%s:%s:%s",
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
pw->pw_name,
pw->pw_passwd ? pw->pw_passwd : "x",
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
(unsigned long)pw->pw_uid,
(unsigned long)pw->pw_gid,
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
pw->pw_gecos,
pw->pw_dir,
pw->pw_shell );
if ( ret < 0 ) {
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
DEBUG(0,("passwd2string: pstr_sprintf() failed!\n"));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return NULL;
}
return string;
}
/**********************************************************************
Convert a string in /etc/group format to a struct group* entry
**********************************************************************/
static WINBINDD_GR* string2group( char *string )
{
static WINBINDD_GR grp;
char *p, *str;
char *fields[NUM_GRP_FIELDS];
int i;
char **gr_members = NULL;
int num_gr_members = 0;
if ( !string )
return NULL;
ZERO_STRUCTP( &grp );
DEBUG(10,("string2group: converting \"%s\"\n", string));
ZERO_STRUCT( fields );
for ( i=0, str=string; i<NUM_GRP_FIELDS-1; i++ ) {
if ( !(p = strchr( str, ':' )) ) {
DEBUG(0,("string2group: parsing failure\n"));
return NULL;
}
*p = '\0';
if ( str )
fields[i] = str;
str = p + 1;
}
/* group members */
if ( *str ) {
/* we already know we have a non-empty string */
num_gr_members = count_chars(str, ',') + 1;
/* if there was at least one comma, then there
are n+1 members */
if ( num_gr_members ) {
fstring buffer;
gr_members = (char**)smb_xmalloc(sizeof(char*)*num_gr_members+1);
i = 0;
while ( next_token(&str, buffer, ",", sizeof(buffer)) && i<num_gr_members ) {
gr_members[i++] = smb_xstrdup(buffer);
}
gr_members[i] = NULL;
}
}
/* copy fields */
fstrcpy( grp.gr_name, fields[0] );
fstrcpy( grp.gr_passwd, fields[1] );
grp.gr_gid = atoi( fields[2] );
grp.num_gr_mem = num_gr_members;
grp.gr_mem = gr_members;
/* last minute sanity checks */
if ( grp.gr_gid == 0 ) {
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(0,("string2group: Failure! gid==%lu\n", (unsigned long)grp.gr_gid));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
SAFE_FREE( gr_members );
return NULL;
}
DEBUG(10,("string2group: Success\n"));
return &grp;
}
/**********************************************************************
Convert a struct group* to a string formatted for /etc/group
**********************************************************************/
static char* group2string( const WINBINDD_GR *grp )
{
static pstring string;
int ret;
char *member, *gr_mem_str;
int num_members;
int i, size;
if ( !grp || !grp->gr_name )
return NULL;
DEBUG(10,("group2string: converting passwd struct for %s\n",
grp->gr_name));
if ( grp->num_gr_mem ) {
int idx = 0;
member = grp->gr_mem[0];
size = 0;
num_members = 0;
while ( member ) {
size += strlen(member) + 1;
num_members++;
member = grp->gr_mem[num_members];
}
gr_mem_str = smb_xmalloc(size);
for ( i=0; i<num_members; i++ ) {
snprintf( &gr_mem_str[idx], size-idx, "%s,", grp->gr_mem[i] );
idx += strlen(grp->gr_mem[i]) + 1;
}
/* add trailing NULL (also removes trailing ',' */
gr_mem_str[size-1] = '\0';
}
else {
/* no members */
gr_mem_str = smb_xmalloc(sizeof(fstring));
fstrcpy( gr_mem_str, "" );
}
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
ret = pstr_sprintf( string, "%s:%s:%lu:%s",
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
grp->gr_name,
grp->gr_passwd ? grp->gr_passwd : "*",
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
(unsigned long)grp->gr_gid,
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
gr_mem_str );
SAFE_FREE( gr_mem_str );
if ( ret < 0 ) {
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
DEBUG(0,("group2string: pstr_sprintf() failed!\n"));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return NULL;
}
return string;
}
/**********************************************************************
**********************************************************************/
static char* acct_userkey_byname( const char *name )
{
static fstring key;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/NAME/%s", WBKEY_PASSWD, name );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return key;
}
/**********************************************************************
**********************************************************************/
static char* acct_userkey_byuid( uid_t uid )
{
static fstring key;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/UID/%lu", WBKEY_PASSWD, (unsigned long)uid );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return key;
}
/**********************************************************************
**********************************************************************/
static char* acct_groupkey_byname( const char *name )
{
static fstring key;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/NAME/%s", WBKEY_GROUP, name );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return key;
}
/**********************************************************************
**********************************************************************/
static char* acct_groupkey_bygid( gid_t gid )
{
static fstring key;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/GID/%lu", WBKEY_GROUP, (unsigned long)gid );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return key;
}
/**********************************************************************
**********************************************************************/
WINBINDD_PW* wb_getpwnam( const char * name )
{
char *keystr;
TDB_DATA data;
static WINBINDD_PW *pw;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_getpwnam: Failed to open winbindd account db\n"));
return NULL;
}
keystr = acct_userkey_byname( name );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, keystr );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
pw = NULL;
if ( data.dptr ) {
pw = string2passwd( data.dptr );
SAFE_FREE( data.dptr );
}
DEBUG(5,("wb_getpwnam: %s user (%s)\n",
(pw ? "Found" : "Did not find"), name ));
return pw;
}
/**********************************************************************
**********************************************************************/
WINBINDD_PW* wb_getpwuid( const uid_t uid )
{
char *keystr;
TDB_DATA data;
static WINBINDD_PW *pw;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_getpwuid: Failed to open winbindd account db\n"));
return NULL;
}
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, acct_userkey_byuid(uid) );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
if ( !data.dptr ) {
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(4,("wb_getpwuid: failed to locate uid == %lu\n", (unsigned long)uid));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return NULL;
}
keystr = acct_userkey_byname( data.dptr );
SAFE_FREE( data.dptr );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, keystr );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
pw = NULL;
if ( data.dptr ) {
pw = string2passwd( data.dptr );
SAFE_FREE( data.dptr );
}
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(5,("wb_getpwuid: %s user (uid == %lu)\n",
(pw ? "Found" : "Did not find"), (unsigned long)uid ));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return pw;
}
/**********************************************************************
**********************************************************************/
BOOL wb_storepwnam( const WINBINDD_PW *pw )
{
char *namekey, *uidkey;
TDB_DATA data;
char *str;
int ret = 0;
fstring username;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_storepwnam: Failed to open winbindd account db\n"));
return False;
}
namekey = acct_userkey_byname( pw->pw_name );
/* lock the main entry first */
if ( tdb_lock_bystring(account_tdb, namekey, 0) == -1 ) {
DEBUG(0,("wb_storepwnam: Failed to lock %s\n", namekey));
return False;
}
str = passwd2string( pw );
data.dptr = str;
data.dsize = strlen(str) + 1;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( (tdb_store_bystring(account_tdb, namekey, data, TDB_REPLACE)) == -1 ) {
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
DEBUG(0,("wb_storepwnam: Failed to store \"%s\"\n", str));
ret = -1;
goto done;
}
/* store the uid index */
uidkey = acct_userkey_byuid(pw->pw_uid);
fstrcpy( username, pw->pw_name );
data.dptr = username;
data.dsize = strlen(username) + 1;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( (tdb_store_bystring(account_tdb, uidkey, data, TDB_REPLACE)) == -1 ) {
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
DEBUG(0,("wb_storepwnam: Failed to store uid key \"%s\"\n", str));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
tdb_delete_bystring(account_tdb, namekey);
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
ret = -1;
goto done;
}
DEBUG(10,("wb_storepwnam: Success -> \"%s\"\n", str));
done:
tdb_unlock_bystring( account_tdb, namekey );
return ( ret == 0 );
}
/**********************************************************************
**********************************************************************/
WINBINDD_GR* wb_getgrnam( const char * name )
{
char *keystr;
TDB_DATA data;
static WINBINDD_GR *grp;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_getgrnam: Failed to open winbindd account db\n"));
return NULL;
}
keystr = acct_groupkey_byname( name );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, keystr );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
grp = NULL;
if ( data.dptr ) {
grp = string2group( data.dptr );
SAFE_FREE( data.dptr );
}
DEBUG(5,("wb_getgrnam: %s group (%s)\n",
(grp ? "Found" : "Did not find"), name ));
return grp;
}
/**********************************************************************
**********************************************************************/
WINBINDD_GR* wb_getgrgid( gid_t gid )
{
char *keystr;
TDB_DATA data;
static WINBINDD_GR *grp;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_getgrgid: Failed to open winbindd account db\n"));
return NULL;
}
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, acct_groupkey_bygid(gid) );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
if ( !data.dptr ) {
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(4,("wb_getgrgid: failed to locate gid == %lu\n",
(unsigned long)gid));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return NULL;
}
keystr = acct_groupkey_byname( data.dptr );
SAFE_FREE( data.dptr );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
data = tdb_fetch_bystring( account_tdb, keystr );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
grp = NULL;
if ( data.dptr ) {
grp = string2group( data.dptr );
SAFE_FREE( data.dptr );
}
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(5,("wb_getgrgid: %s group (gid == %lu)\n",
(grp ? "Found" : "Did not find"), (unsigned long)gid ));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return grp;
}
/**********************************************************************
**********************************************************************/
BOOL wb_storegrnam( const WINBINDD_GR *grp )
{
char *namekey, *gidkey;
TDB_DATA data;
char *str;
int ret = 0;
fstring groupname;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_storepwnam: Failed to open winbindd account db\n"));
return False;
}
namekey = acct_groupkey_byname( grp->gr_name );
/* lock the main entry first */
if ( tdb_lock_bystring(account_tdb, namekey, 0) == -1 ) {
DEBUG(0,("wb_storegrnam: Failed to lock %s\n", namekey));
return False;
}
str = group2string( grp );
data.dptr = str;
data.dsize = strlen(str) + 1;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( (tdb_store_bystring(account_tdb, namekey, data, TDB_REPLACE)) == -1 ) {
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
DEBUG(0,("wb_storegrnam: Failed to store \"%s\"\n", str));
ret = -1;
goto done;
}
/* store the gid index */
gidkey = acct_groupkey_bygid(grp->gr_gid);
fstrcpy( groupname, grp->gr_name );
data.dptr = groupname;
data.dsize = strlen(groupname) + 1;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( (tdb_store_bystring(account_tdb, gidkey, data, TDB_REPLACE)) == -1 ) {
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
DEBUG(0,("wb_storegrnam: Failed to store gid key \"%s\"\n", str));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
tdb_delete_bystring(account_tdb, namekey);
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
ret = -1;
goto done;
}
DEBUG(10,("wb_storegrnam: Success -> \"%s\"\n", str));
done:
tdb_unlock_bystring( account_tdb, namekey );
return ( ret == 0 );
}
/**********************************************************************
**********************************************************************/
static BOOL wb_addgrpmember( WINBINDD_GR *grp, const char *user )
{
int i;
char **members;
if ( !grp || !user )
return False;
for ( i=0; i<grp->num_gr_mem; i++ ) {
if ( StrCaseCmp( grp->gr_mem[i], user ) == 0 )
return True;
}
/* add one new slot and keep an extra for the terminating NULL */
members = Realloc( grp->gr_mem, (grp->num_gr_mem+2)*sizeof(char*) );
if ( !members )
return False;
grp->gr_mem = members;
grp->gr_mem[grp->num_gr_mem++] = smb_xstrdup(user);
grp->gr_mem[grp->num_gr_mem] = NULL;
return True;
}
/**********************************************************************
**********************************************************************/
static BOOL wb_delgrpmember( WINBINDD_GR *grp, const char *user )
{
int i;
BOOL found = False;
if ( !grp || !user )
return False;
for ( i=0; i<grp->num_gr_mem && !found; i++ ) {
if ( StrCaseCmp( grp->gr_mem[i], user ) == 0 )
found = True;
}
if ( !found )
return False;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
/* still some remaining members */
if ( grp->num_gr_mem > 1 ) {
memmove( grp->gr_mem[i], grp->gr_mem[i+1], sizeof(char*)*(grp->num_gr_mem-(i+1)) );
grp->num_gr_mem--;
}
else { /* last one */
free_winbindd_gr( grp );
grp->gr_mem = NULL;
grp->num_gr_mem = 0;
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
return True;
}
/**********************************************************************
**********************************************************************/
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
static int cleangroups_traverse_fn(TDB_CONTEXT *the_tdb, TDB_DATA kbuf, TDB_DATA dbuf,
void *state)
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
{
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
int len;
fstring key;
char *name = (char*)state;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/NAME", WBKEY_GROUP );
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
len = strlen(key);
/* if this is a group entry then, check the members */
if ( (strncmp(kbuf.dptr, key, len) == 0) && dbuf.dptr ) {
WINBINDD_GR *grp;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
if ( !(grp = string2group( dbuf.dptr )) ) {
DEBUG(0,("cleangroups_traverse_fn: Failure to parse [%s]\n",
dbuf.dptr));
return 0;
}
/* just try to delete the user and rely on wb_delgrpmember()
to tell you whether or not the group changed. This is more
effecient than testing group membership first since the
checks for deleting a user from a group is essentially the
same as checking if he/she is a member */
if ( wb_delgrpmember( grp, name ) ) {
DEBUG(10,("cleanupgroups_traverse_fn: Removed user (%s) from group (%s)\n",
name, grp->gr_name));
wb_storegrnam( grp );
}
free_winbindd_gr( grp );
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
return 0;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
/**********************************************************************
**********************************************************************/
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
static BOOL wb_delete_user( WINBINDD_PW *pw)
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
{
char *namekey;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
char *uidkey;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( !account_tdb && !winbindd_accountdb_init() ) {
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
DEBUG(0,("wb_delete_user: Failed to open winbindd account db\n"));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
return False;
}
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
namekey = acct_userkey_byname( pw->pw_name );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
/* lock the main entry first */
if ( tdb_lock_bystring(account_tdb, namekey, 0) == -1 ) {
DEBUG(0,("wb_delete_user: Failed to lock %s\n", namekey));
return False;
}
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
/* remove user from all groups */
tdb_traverse(account_tdb, cleangroups_traverse_fn, (void *)pw->pw_name);
/* remove the user */
uidkey = acct_userkey_byuid( pw->pw_uid );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
tdb_delete_bystring( account_tdb, namekey );
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
tdb_delete_bystring( account_tdb, uidkey );
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
tdb_unlock_bystring( account_tdb, namekey );
return True;
}
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
/**********************************************************************
**********************************************************************/
static int isprimarygroup_traverse_fn(TDB_CONTEXT *the_tdb, TDB_DATA kbuf,
TDB_DATA dbuf, void *params)
{
int len;
fstring key;
struct _check_primary_grp *check = (struct _check_primary_grp*)params;
convert snprintf() calls using pstrings & fstrings to pstr_sprintf() and fstr_sprintf() to try to standardize. lots of snprintf() calls were using len-1; some were using len. At least this helps to be consistent. (This used to be commit 9f835b85dd38cbe655eb19021ff763f31886ac00)
2003-07-23 12:33:59 +00:00
fstr_sprintf( key, "%s/NAME", WBKEY_PASSWD );
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
len = strlen(key);
/* if this is a group entry then, check the members */
if ( (strncmp(kbuf.dptr, key, len) == 0) && dbuf.dptr ) {
WINBINDD_PW *pw;;
if ( !(pw = string2passwd( dbuf.dptr )) ) {
DEBUG(0,("isprimarygroup_traverse_fn: Failure to parse [%s]\n",
dbuf.dptr));
return 0;
}
if ( check->gid == pw->pw_gid ) {
check->found = True;
return 1;
}
}
return 0;
}
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
/**********************************************************************
**********************************************************************/
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
static BOOL wb_delete_group( WINBINDD_GR *grp )
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
{
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
struct _check_primary_grp check;
char *namekey;
char *gidkey;
if ( !account_tdb && !winbindd_accountdb_init() ) {
DEBUG(0,("wb_delete_group: Failed to open winbindd account db\n"));
return False;
}
/* lock the main entry first */
namekey = acct_groupkey_byname( grp->gr_name );
if ( tdb_lock_bystring(account_tdb, namekey, 0) == -1 ) {
DEBUG(0,("wb_delete_group: Failed to lock %s\n", namekey));
return False;
}
/* is this group the primary group for any user? If
so deny delete */
check.found = False;
tdb_traverse(account_tdb, isprimarygroup_traverse_fn, (void *)&check);
if ( check.found ) {
DEBUG(4,("wb_delete_group: Cannot delete group (%s) since it "
"is the primary group for some users\n", grp->gr_name));
return False;
}
/* We're clear. Delete the group */
DEBUG(5,("wb_delete_group: Removing group (%s)\n", grp->gr_name));
gidkey = acct_groupkey_bygid( grp->gr_gid );
tdb_delete_bystring( account_tdb, namekey );
tdb_delete_bystring( account_tdb, gidkey );
tdb_unlock_bystring( account_tdb, namekey );
return True;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
/**********************************************************************
Create a new "UNIX" user for the system given a username
**********************************************************************/
enum winbindd_result winbindd_create_user(struct winbindd_cli_state *state)
{
char *user, *group;
unid_t id;
WINBINDD_PW pw;
WINBINDD_GR *wb_grp;
struct group *unix_grp;
gid_t primary_gid;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
uint32 flags = state->request.flags;
uint32 rid;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
if ( !state->privileged ) {
DEBUG(2, ("winbindd_create_user: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.username)-1]='\0';
state->request.data.acct_mgt.groupname[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
user = state->request.data.acct_mgt.username;
group = state->request.data.acct_mgt.groupname;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: create_user: user=>(%s), group=>(%s)\n",
(unsigned long)state->pid, user, group));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
if ( !*group )
group = lp_template_primary_group();
/* validate the primary group
1) lookup in local tdb first
2) call getgrnam() as a last resort */
if ( (wb_grp=wb_getgrnam(group)) != NULL ) {
primary_gid = wb_grp->gr_gid;
free_winbindd_gr( wb_grp );
}
else if ( (unix_grp=sys_getgrnam(group)) != NULL ) {
primary_gid = unix_grp->gr_gid;
}
else {
DEBUG(2,("winbindd_create_user: Cannot validate gid for group (%s)\n", group));
return WINBINDD_ERROR;
}
/* get a new uid */
if ( !NT_STATUS_IS_OK(idmap_allocate_id( &id, ID_USERID)) ) {
DEBUG(0,("winbindd_create_user: idmap_allocate_id() failed!\n"));
return WINBINDD_ERROR;
}
/* The substitution of %U and %D in the 'template homedir' is done
by lp_string() calling standard_sub_basic(). */
fstrcpy( current_user_info.smb_name, user );
sub_set_smb_name( user );
fstrcpy( current_user_info.domain, get_global_sam_name() );
/* fill in the passwd struct */
fstrcpy( pw.pw_name, user );
fstrcpy( pw.pw_passwd, "x" );
fstrcpy( pw.pw_gecos, user);
fstrcpy( pw.pw_dir, lp_template_homedir() );
fstrcpy( pw.pw_shell, lp_template_shell() );
pw.pw_uid = id.uid;
pw.pw_gid = primary_gid;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
/* store the new entry */
if ( !wb_storepwnam(&pw) )
return WINBINDD_ERROR;
/* do we need a new RID? */
if ( flags & WBFLAG_ALLOCATE_RID ) {
if ( !NT_STATUS_IS_OK(idmap_allocate_rid(&rid, USER_RID_TYPE)) ) {
DEBUG(0,("winbindd_create_user: RID allocation failure! Cannot create user (%s)\n",
user));
wb_delete_user( &pw );
return WINBINDD_ERROR;
}
state->response.data.rid = rid;
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
return WINBINDD_OK;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
/**********************************************************************
Create a new "UNIX" group for the system given a username
**********************************************************************/
enum winbindd_result winbindd_create_group(struct winbindd_cli_state *state)
{
char *group;
unid_t id;
WINBINDD_GR grp;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
uint32 flags = state->request.flags;
uint32 rid;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
if ( !state->privileged ) {
DEBUG(2, ("winbindd_create_group: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.groupname[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
group = state->request.data.acct_mgt.groupname;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: create_group: (%s)\n", (unsigned long)state->pid, group));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
/* get a new uid */
if ( !NT_STATUS_IS_OK(idmap_allocate_id( &id, ID_GROUPID)) ) {
DEBUG(0,("winbindd_create_group: idmap_allocate_id() failed!\n"));
return WINBINDD_ERROR;
}
/* fill in the group struct */
fstrcpy( grp.gr_name, group );
fstrcpy( grp.gr_passwd, "*" );
grp.gr_gid = id.gid;
grp.gr_mem = NULL; /* start with no members */
grp.num_gr_mem = 0;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
if ( !wb_storegrnam(&grp) )
return WINBINDD_ERROR;
/* do we need a new RID? */
if ( flags & WBFLAG_ALLOCATE_RID ) {
if ( !NT_STATUS_IS_OK(idmap_allocate_rid(&rid, GROUP_RID_TYPE)) ) {
DEBUG(0,("winbindd_create_group: RID allocation failure! Cannot create group (%s)\n",
group));
wb_delete_group( &grp );
return WINBINDD_ERROR;
}
state->response.data.rid = rid;
}
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
return WINBINDD_OK;
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
/**********************************************************************
Add a user to the membership for a group.
**********************************************************************/
enum winbindd_result winbindd_add_user_to_group(struct winbindd_cli_state *state)
{
WINBINDD_PW *pw;
WINBINDD_GR *grp;
char *user, *group;
BOOL ret;
if ( !state->privileged ) {
DEBUG(2, ("winbindd_add_user_to_group: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.groupname[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.username)-1]='\0';
group = state->request.data.acct_mgt.groupname;
user = state->request.data.acct_mgt.username;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: add_user_to_group: add %s to %s\n", (unsigned long)state->pid,
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
user, group));
/* make sure it is a valid user */
if ( !(pw = wb_getpwnam( user )) ) {
DEBUG(4,("winbindd_add_user_to_group: Cannot add a non-existent user\n"));
return WINBINDD_ERROR;
}
/* make sure it is a valid group */
if ( !(grp = wb_getgrnam( group )) ) {
DEBUG(4,("winbindd_add_user_to_group: Cannot add a user to a non-extistent group\n"));
return WINBINDD_ERROR;
}
if ( !wb_addgrpmember( grp, user ) )
return WINBINDD_ERROR;
ret = wb_storegrnam(grp);
free_winbindd_gr( grp );
return ( ret ? WINBINDD_OK : WINBINDD_ERROR );
}
/**********************************************************************
Remove a user from the membership of a group
**********************************************************************/
enum winbindd_result winbindd_remove_user_from_group(struct winbindd_cli_state *state)
{
WINBINDD_GR *grp;
char *user, *group;
BOOL ret;
if ( !state->privileged ) {
DEBUG(2, ("winbindd_remove_user_from_group: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.groupname[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.username)-1]='\0';
group = state->request.data.acct_mgt.groupname;
user = state->request.data.acct_mgt.username;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: remove_user_to_group: delete %s from %s\n", (unsigned long)state->pid,
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
user, group));
/* don't worry about checking the username since we're removing it anyways */
/* make sure it is a valid group */
if ( !(grp = wb_getgrnam( group )) ) {
DEBUG(4,("winbindd_remove_user_to_group: Cannot remove a user to a non-extistent group\n"));
return WINBINDD_ERROR;
}
if ( !wb_delgrpmember( grp, user ) )
return WINBINDD_ERROR;
ret = wb_storegrnam(grp);
free_winbindd_gr( grp );
return ( ret ? WINBINDD_OK : WINBINDD_ERROR );
}
/**********************************************************************
Set the primary group membership of a user
**********************************************************************/
enum winbindd_result winbindd_set_user_primary_group(struct winbindd_cli_state *state)
{
WINBINDD_PW *pw;
WINBINDD_GR *grp;
char *user, *group;
if ( !state->privileged ) {
DEBUG(2, ("winbindd_set_user_primary_group: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.groupname[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.username)-1]='\0';
group = state->request.data.acct_mgt.groupname;
user = state->request.data.acct_mgt.username;
Another round of uid/gid/pid format string changes I missed the first time. (This used to be commit 6616485dbad74dab7506609c6bfd183fc9c1f93c)
2003-07-22 06:52:39 +00:00
DEBUG(3, ("[%5lu]: set_user_primary_group: group %s for user %s\n",
(unsigned long)state->pid, group, user));
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
/* make sure it is a valid user */
if ( !(pw = wb_getpwnam( user )) ) {
DEBUG(4,("winbindd_add_user_to_group: Cannot add a non-existent user\n"));
return WINBINDD_ERROR;
}
/* make sure it is a valid group */
if ( !(grp = wb_getgrnam( group )) ) {
DEBUG(4,("winbindd_add_user_to_group: Cannot add a user to a non-extistent group\n"));
return WINBINDD_ERROR;
}
pw->pw_gid = grp->gr_gid;
free_winbindd_gr( grp );
return ( wb_storepwnam(pw) ? WINBINDD_OK : WINBINDD_ERROR );
}
/**********************************************************************
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
Delete a user from the winbindd account tdb.
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
**********************************************************************/
enum winbindd_result winbindd_delete_user(struct winbindd_cli_state *state)
{
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
WINBINDD_PW *pw;
char *user;
if ( !state->privileged ) {
DEBUG(2, ("winbindd_delete_user: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.username)-1]='\0';
user = state->request.data.acct_mgt.username;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: delete_user: %s\n", (unsigned long)state->pid, user));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
/* make sure it is a valid user */
if ( !(pw = wb_getpwnam( user )) ) {
DEBUG(4,("winbindd_delete_user: Cannot delete a non-existent user\n"));
return WINBINDD_ERROR;
}
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
return ( wb_delete_user(pw) ? WINBINDD_OK : WINBINDD_ERROR );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
/**********************************************************************
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
Delete a group from winbindd's account tdb.
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
**********************************************************************/
enum winbindd_result winbindd_delete_group(struct winbindd_cli_state *state)
{
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
WINBINDD_GR *grp;
char *group;
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
BOOL ret;
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
if ( !state->privileged ) {
DEBUG(2, ("winbindd_delete_group: non-privileged access denied!\n"));
return WINBINDD_ERROR;
}
/* Ensure null termination */
state->request.data.acct_mgt.username[sizeof(state->request.data.acct_mgt.groupname)-1]='\0';
group = state->request.data.acct_mgt.groupname;
Fixup a bunch of printf-style functions and debugs to use unsigned long when displaying pid_t, uid_t and gid_t values. This removes a whole lot of warnings on some of the 64-bit build farm machines as well as help us out when 64-bit uid/gid/pid values come along. (This used to be commit f93528ba007c8800a850678f35f499fb7360fb9a)
2003-07-22 04:31:20 +00:00
DEBUG(3, ("[%5lu]: delete_group: %s\n", (unsigned long)state->pid, group));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
/* make sure it is a valid group */
if ( !(grp = wb_getgrnam( group )) ) {
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
DEBUG(4,("winbindd_delete_group: Cannot delete a non-existent group\n"));
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
return WINBINDD_ERROR;
}
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
ret = wb_delete_group(grp);
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
free_winbindd_gr( grp );
moving more code around. * move rid allocation into IDMAP. See comments in _api_samr_create_user() * add winbind delete user/group functions I'm checking this in to sync up with everyone. But I'm going to split the add a separate winbindd_allocate_rid() function for systems that have an 'add user script' but need idmap to give them a RID. Life would be so much simplier without 'enable rid algorithm'. The current RID allocation is horrible due to this one fact. Tested idmap_tdb but not idmap_ldap yet. Will do that tomorrow. Nothing has changed in the way a samba domain is represented, stored, or search in the directory so things should be ok with previous installations. going to bed now. (This used to be commit 0463045cc7ff177fab44b25faffad5bf7140244d)
2003-07-11 05:33:40 +00:00
return ( ret ? WINBINDD_OK : WINBINDD_ERROR );
Large set of changes to add UNIX account/group management to winbindd. See README.idmap-and-winbind-changes for details. (This used to be commit 1111bc7b0c7165e1cdf8d90eb49f4c368d2eded6)
2003-07-09 16:44:47 +00:00
}
i guess i'm the only one this ever annyoed... fix the confusion when we tdb_lock_bystring() but we retrieve an entry using tdb_fetch_by_string. It's now always tdb.*bystring() (This used to be commit 66359531b89368939f0e8f584a45844b5f2f99e7)
2003-07-10 20:37:01 +00:00
Copy Permalink