sin/samba-mirror
1
0
Fork 0
mirror of https://github.com/samba-team/samba.git synced 2024-12-24 21:34:56 +03:00
Code
2cddd5fe8c
samba-mirror/source3/smbd/service.c

702 lines
18 KiB
C
Raw Normal View History

this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/*
Unix SMB/Netbios implementation.
Version 1.9.
service (connection) opening and closing
Copyright (C) Andrew Tridgell 1992-1998
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 "includes.h"
extern int DEBUGLEVEL;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
extern struct timeval smb_last_time;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
extern int case_default;
extern BOOL case_preserve;
extern BOOL short_case_preserve;
extern BOOL case_mangle;
extern BOOL case_sensitive;
extern BOOL use_mangled_map;
extern fstring remote_machine;
include/vfs.h: smbd/vfs-wrap.c: smbd/vfs.c: Added fchmod_acl and chmod_acl. lib/substitute.c: smbd/lanman.c: smbd/open.c: smbd/process.c: smbd/reply.c: smbd/service.c: Removed sessetup_user variable. Added current_user_info struct which conatins domain info etc. Added '%D' for client domain parameter. Jeremy. (This used to be commit 2844ec3d511680609d6794b8718001a1bda9e89f)
2001-01-23 04:52:30 +03:00
extern userdom_struct current_user_info;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
extern fstring remote_machine;
/****************************************************************************
load parameters specific to a connection/service
****************************************************************************/
BOOL become_service(connection_struct *conn,BOOL do_chdir)
{
extern char magic_char;
static connection_struct *last_conn;
int snum;
if (!conn) {
last_conn = NULL;
return(False);
}
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
conn->lastused = smb_last_time.tv_sec;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
snum = SNUM(conn);
if (do_chdir &&
Restructuring of the code to remove dos_ChDir/dos_GetWd and re-vector them through the VFS. All file access/directory access code in smbd should now go via the vfs. Added vfs_chown/vfs_chmod calls. Still looking at vfs_get_nt_acl() vfs_set_nt_acl() call API design. Jeremy. (This used to be commit f96625ec124adb6e110dc54632e006b3620a962b)
2000-09-27 23:09:59 +04:00
vfs_ChDir(conn,conn->connectpath) != 0 &&
vfs_ChDir(conn,conn->origpath) != 0) {
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
DEBUG(0,("chdir (%s) failed\n",
conn->connectpath));
return(False);
}
if (conn == last_conn)
return(True);
last_conn = conn;
case_default = lp_defaultcase(snum);
case_preserve = lp_preservecase(snum);
short_case_preserve = lp_shortpreservecase(snum);
case_mangle = lp_casemangle(snum);
case_sensitive = lp_casesensitive(snum);
magic_char = lp_magicchar(snum);
use_mangled_map = (*lp_mangled_map(snum) ? True:False);
return(True);
}
smbd/process.c: & type with 0xff for paranioa sake... smbd/reply.c smbd/service.c: cause all "add home service" calls to go through a winbindd aware function. Jeremy. (This used to be commit a72d12e992e2755e925032aef1aa99be74bf6652)
2001-01-24 22:34:53 +03:00
/****************************************************************************
Add a home service. Returns the new service number or -1 if fail.
****************************************************************************/
int add_home_service(char *service, char *homedir)
{
int iHomeService;
int iService;
fstring new_service;
char *usr_p = NULL;
if (!service || !homedir)
return -1;
if ((iHomeService = lp_servicenumber(HOMES_NAME)) < 0)
return -1;
/*
* If this is a winbindd provided username, remove
* the domain component before adding the service.
* Log a warning if the "path=" parameter does not
* include any macros.
*/
fstrcpy(new_service, service);
if ((usr_p = strchr(service,*lp_winbind_separator())) != NULL)
fstrcpy(new_service, usr_p+1);
lp_add_home(new_service,iHomeService,homedir);
iService = lp_servicenumber(new_service);
return iService;
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/****************************************************************************
smbd/process.c: & type with 0xff for paranioa sake... smbd/reply.c smbd/service.c: cause all "add home service" calls to go through a winbindd aware function. Jeremy. (This used to be commit a72d12e992e2755e925032aef1aa99be74bf6652)
2001-01-24 22:34:53 +03:00
Find a service entry. service is always in dos codepage.
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
****************************************************************************/
smbd/process.c: & type with 0xff for paranioa sake... smbd/reply.c smbd/service.c: cause all "add home service" calls to go through a winbindd aware function. Jeremy. (This used to be commit a72d12e992e2755e925032aef1aa99be74bf6652)
2001-01-24 22:34:53 +03:00
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
int find_service(char *service)
{
int iService;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
all_string_sub(service,"\\","/",0);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
iService = lp_servicenumber(service);
/* now handle the special case of a home directory */
if (iService < 0)
{
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
char *phome_dir = get_user_home_dir(service);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
if(!phome_dir)
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
{
/*
* Try mapping the servicename, it may
* be a Windows to unix mapped user name.
*/
if(map_username(service))
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
phome_dir = get_user_home_dir(service);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
DEBUG(3,("checking for home directory %s gave %s\n",service,
phome_dir?phome_dir:"(NULL)"));
smbd/process.c: & type with 0xff for paranioa sake... smbd/reply.c smbd/service.c: cause all "add home service" calls to go through a winbindd aware function. Jeremy. (This used to be commit a72d12e992e2755e925032aef1aa99be74bf6652)
2001-01-24 22:34:53 +03:00
iService = add_home_service(service,phome_dir);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
/* If we still don't have a service, attempt to add it as a printer. */
if (iService < 0)
{
int iPrinterService;
if ((iPrinterService = lp_servicenumber(PRINTERS_NAME)) >= 0)
{
char *pszTemp;
DEBUG(3,("checking whether %s is a valid printer name...\n", service));
pszTemp = PRINTCAP;
if ((pszTemp != NULL) && pcap_printername_ok(service, pszTemp))
{
DEBUG(3,("%s is a valid printer name\n", service));
DEBUG(3,("adding %s as a printer service\n", service));
lp_add_printer(service,iPrinterService);
iService = lp_servicenumber(service);
if (iService < 0)
DEBUG(0,("failed to add %s as a printer service!\n", service));
}
else
DEBUG(3,("%s is not a valid printer name\n", service));
}
}
Mega-VFS merge. Yeah baby! Synopsis: change every disk access function to work through a vfs_ops structure contained in the connection_struct. (This used to be commit 3aad500c0fb61232ed3431ff4b743b5d18ec852f)
2000-02-03 08:17:25 +03:00
/* Check for default vfs service? Unsure whether to implement this */
if (iService < 0)
{
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/* just possibly it's a default service? */
if (iService < 0)
{
char *pdefservice = lp_defaultservice();
don't allow ".." in service name when doing "default service" processing. (This used to be commit 702263bba555a1d7c7999d40e5789b7e920dbce4)
1998-11-29 09:23:16 +03:00
if (pdefservice && *pdefservice &&
!strequal(pdefservice,service) &&
!strstr(service,".."))
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
{
/*
* We need to do a local copy here as lp_defaultservice()
* returns one of the rotating lp_string buffers that
* could get overwritten by the recursive find_service() call
* below. Fix from Josef Hinteregger <joehtg@joehtg.co.at>.
*/
pstring defservice;
pstrcpy(defservice, pdefservice);
iService = find_service(defservice);
if (iService >= 0)
{
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
all_string_sub(service,"_","/",0);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
iService = lp_add_service(service,iService);
}
}
}
if (iService >= 0)
if (!VALID_SNUM(iService))
{
DEBUG(0,("Invalid snum %d for %s\n",iService,service));
iService = -1;
}
if (iService < 0)
DEBUG(3,("find_service() failed to find service %s\n", service));
return (iService);
}
/****************************************************************************
make a connection to a service
****************************************************************************/
connection_struct *make_connection(char *service,char *user,char *password, int pwlen, char *dev,uint16 vuid, int *ecode)
{
int snum;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
struct passwd *pass = NULL;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
BOOL guest = False;
BOOL force = False;
connection_struct *conn;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
int ret;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
strlower(service);
snum = find_service(service);
if (snum < 0) {
Added ADMIN$ share as alias for IPC$ except no guest connect. AS/U wants to do RPC calls down this treeid. Jeremy. (This used to be commit 83133bab0ed59e303a183fd91812165f08e88484)
2001-03-13 03:31:07 +03:00
if (strequal(service,"IPC$") || strequal(service,"ADMIN$")) {
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
DEBUG(3,("refusing IPC connection\n"));
*ecode = ERRnoipc;
return NULL;
}
DEBUG(0,("%s (%s) couldn't find service %s\n",
finally got sick of the "extern int Client" code and the stupid assumption that we have one socket everywhere while doing so I discovered a few bugs! 1) the clientgen session retarget code if used from smbd or nmbd would cause a crash as it called close_sockets() which closed our main socket! fixed by removing close_sockets() completely - it is unnecessary 2) the caching in client_addr() and client_name() was bogus - it could easily get fooled and give the wrong result. fixed. 3) the retarget could could recurse, allowing an easy denial of service attack on nmbd. fixed. (This used to be commit 5937ab14d222696e40a3fc6f0e6a536f2d7305d3)
2000-04-11 17:55:53 +04:00
remote_machine, client_addr(), service));
Fixed bug where file access was allowed on IPC$ share. Return correct error codes on invalid share name. Jeremy. (This used to be commit 420d6bc4809cef9d74354175d0fa956ab4e8ac3c)
2000-05-23 05:27:19 +04:00
*ecode = ERRnosuchshare;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
return NULL;
}
if (strequal(service,HOMES_NAME)) {
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
if (*user && Get_Pwnam(user,True)) {
fstring dos_username;
fstrcpy(dos_username, user);
unix_to_dos(dos_username, True);
return(make_connection(dos_username,user,password,
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
pwlen,dev,vuid,ecode));
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
if(lp_security() != SEC_SHARE) {
if (validated_username(vuid)) {
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
fstring dos_username;
fstrcpy(user,validated_username(vuid));
fstrcpy(dos_username, user);
unix_to_dos(dos_username, True);
return(make_connection(dos_username,user,password,pwlen,dev,vuid,ecode));
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
} else {
include/vfs.h: smbd/vfs-wrap.c: smbd/vfs.c: Added fchmod_acl and chmod_acl. lib/substitute.c: smbd/lanman.c: smbd/open.c: smbd/process.c: smbd/reply.c: smbd/service.c: Removed sessetup_user variable. Added current_user_info struct which conatins domain info etc. Added '%D' for client domain parameter. Jeremy. (This used to be commit 2844ec3d511680609d6794b8718001a1bda9e89f)
2001-01-23 04:52:30 +03:00
/* Security = share. Try with current_user_info.smb_name
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
* as the username. */
include/vfs.h: smbd/vfs-wrap.c: smbd/vfs.c: Added fchmod_acl and chmod_acl. lib/substitute.c: smbd/lanman.c: smbd/open.c: smbd/process.c: smbd/reply.c: smbd/service.c: Removed sessetup_user variable. Added current_user_info struct which conatins domain info etc. Added '%D' for client domain parameter. Jeremy. (This used to be commit 2844ec3d511680609d6794b8718001a1bda9e89f)
2001-01-23 04:52:30 +03:00
if(*current_user_info.smb_name) {
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
fstring dos_username;
include/vfs.h: smbd/vfs-wrap.c: smbd/vfs.c: Added fchmod_acl and chmod_acl. lib/substitute.c: smbd/lanman.c: smbd/open.c: smbd/process.c: smbd/reply.c: smbd/service.c: Removed sessetup_user variable. Added current_user_info struct which conatins domain info etc. Added '%D' for client domain parameter. Jeremy. (This used to be commit 2844ec3d511680609d6794b8718001a1bda9e89f)
2001-01-23 04:52:30 +03:00
fstrcpy(user,current_user_info.smb_name);
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
fstrcpy(dos_username, user);
unix_to_dos(dos_username, True);
return(make_connection(dos_username,user,password,pwlen,dev,vuid,ecode));
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
}
}
if (!lp_snum_ok(snum) ||
finally got sick of the "extern int Client" code and the stupid assumption that we have one socket everywhere while doing so I discovered a few bugs! 1) the clientgen session retarget code if used from smbd or nmbd would cause a crash as it called close_sockets() which closed our main socket! fixed by removing close_sockets() completely - it is unnecessary 2) the caching in client_addr() and client_name() was bogus - it could easily get fooled and give the wrong result. fixed. 3) the retarget could could recurse, allowing an easy denial of service attack on nmbd. fixed. (This used to be commit 5937ab14d222696e40a3fc6f0e6a536f2d7305d3)
2000-04-11 17:55:53 +04:00
!check_access(smbd_server_fd(),
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
lp_hostsallow(snum), lp_hostsdeny(snum))) {
*ecode = ERRaccess;
return NULL;
}
/* you can only connect to the IPC$ service as an ipc device */
Added ADMIN$ share as alias for IPC$ except no guest connect. AS/U wants to do RPC calls down this treeid. Jeremy. (This used to be commit 83133bab0ed59e303a183fd91812165f08e88484)
2001-03-13 03:31:07 +03:00
if (strequal(service,"IPC$") || strequal(service,"ADMIN$"))
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
pstrcpy(dev,"IPC");
if (*dev == '?' || !*dev) {
if (lp_print_ok(snum)) {
pstrcpy(dev,"LPT1:");
} else {
pstrcpy(dev,"A:");
}
}
/* if the request is as a printer and you can't print then refuse */
strupper(dev);
if (!lp_print_ok(snum) && (strncmp(dev,"LPT",3) == 0)) {
DEBUG(1,("Attempt to connect to non-printer as a printer\n"));
*ecode = ERRinvdevice;
return NULL;
}
Fix for mounting a printer as a share. Not that there's anything specifically wrong with this, but Samba is fooled by the client into thinking the printer is a file share. Files copied to the share gather dust in the spool directory and aren't printed. This patch has the effect of not allowing printers to be mounted as file shares. Not sure whether this is the correct solution or not. {Jeremy,JF,Tridge} please check! (This used to be commit dcf3249bb9fef2a05b376e9c8c1a0a7d602d8a2e)
2000-07-25 10:10:59 +04:00
/* Behave as a printer if we are supposed to */
if (lp_print_ok(snum) && (strcmp(dev, "A:") == 0)) {
pstrcpy(dev, "LPT1:");
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/* lowercase the user name */
strlower(user);
only add the service name and client machine name to list of users names for a session when in share mode security --jerry (This used to be commit 22d6c2c163dd578365bff85ef95abfa59fe356ea)
2001-01-24 01:13:41 +03:00
/* add it as a possible user name if we
are in share mode security */
if (lp_security() == SEC_SHARE) {
add_session_user(service);
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/* shall we let them in? */
if (!authorise_login(snum,user,password,pwlen,&guest,&force,vuid)) {
added spool_io_printer_driver_info_level_6() thsi function and the associated header structure were autogenerated using a little awk based code geerator I wroe ths evening. I'll commit that next ... (This used to be commit 974813f0d4afb6c14ed27c48ab24b19932557f9f)
2000-05-12 18:28:46 +04:00
DEBUG( 2, ( "Invalid username/password for %s [%s]\n", service, user ) );
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
*ecode = ERRbadpw;
return NULL;
}
conn = conn_new();
if (!conn) {
DEBUG(0,("Couldn't find free connection.\n"));
*ecode = ERRnoresource;
return NULL;
}
/* find out some info about the user */
- changed smb_getpwnam() to use winbind style usernames - finished ntdom -> winbind rename in head (This used to be commit ada483cb56453afc6df4ec4be18bfe5e943c7150)
2000-05-10 18:48:33 +04:00
pass = smb_getpwnam(user,True);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
if (pass == NULL) {
DEBUG(0,( "Couldn't find account %s\n",user));
*ecode = ERRbaduid;
conn_free(conn);
return NULL;
}
conn->read_only = lp_readonly(snum);
Added set/get SD's on shares. Check before tcon. Jeremy. (This used to be commit 036b1a8b09fe6a7cca83d631624145574acad7f2)
2001-04-09 10:36:38 +04:00
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
{
pstring list;
StrnCpy(list,lp_readlist(snum),sizeof(pstring)-1);
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
pstring_sub(list,"%S",service);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
if (user_in_list(user,list))
conn->read_only = True;
StrnCpy(list,lp_writelist(snum),sizeof(pstring)-1);
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
pstring_sub(list,"%S",service);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
if (user_in_list(user,list))
conn->read_only = False;
}
/* admin user check */
/* JRA - original code denied admin user if the share was
marked read_only. Changed as I don't think this is needed,
but old code left in case there is a problem here.
*/
if (user_in_list(user,lp_admin_users(snum))
#if 0
&& !conn->read_only
#endif
) {
conn->admin_user = True;
DEBUG(0,("%s logged in as admin user (root privileges)\n",user));
} else {
conn->admin_user = False;
}
conn->force_user = force;
conn->vuid = vuid;
conn->uid = pass->pw_uid;
conn->gid = pass->pw_gid;
finally got sick of the "extern int Client" code and the stupid assumption that we have one socket everywhere while doing so I discovered a few bugs! 1) the clientgen session retarget code if used from smbd or nmbd would cause a crash as it called close_sockets() which closed our main socket! fixed by removing close_sockets() completely - it is unnecessary 2) the caching in client_addr() and client_name() was bogus - it could easily get fooled and give the wrong result. fixed. 3) the retarget could could recurse, allowing an easy denial of service attack on nmbd. fixed. (This used to be commit 5937ab14d222696e40a3fc6f0e6a536f2d7305d3)
2000-04-11 17:55:53 +04:00
safe_strcpy(conn->client_address, client_addr(), sizeof(conn->client_address)-1);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
conn->num_files_open = 0;
conn->lastused = time(NULL);
conn->service = snum;
conn->used = True;
conn->printer = (strncmp(dev,"LPT",3) == 0);
Added ADMIN$ share as alias for IPC$ except no guest connect. AS/U wants to do RPC calls down this treeid. Jeremy. (This used to be commit 83133bab0ed59e303a183fd91812165f08e88484)
2001-03-13 03:31:07 +03:00
conn->ipc = ((strncmp(dev,"IPC",3) == 0) || strequal(dev,"ADMIN$"));
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
conn->dirptr = NULL;
conn->veto_list = NULL;
conn->hide_list = NULL;
conn->veto_oplock_list = NULL;
string_set(&conn->dirpath,"");
string_set(&conn->user,user);
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
conn->nt_user_token = NULL;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
/*
* If force user is true, then store the
* given userid and also the primary groupid
* of the user we're forcing.
*/
if (*lp_force_user(snum)) {
struct passwd *pass2;
pstring fuser;
pstrcpy(fuser,lp_force_user(snum));
Call VFS initialisation, share connect and share disconnect functions. Initialise VFS function pointers with symbols from the 'vfs object' file, use disk pointers otherwise. Added a hook to check for a 'default' VFS share (like the [homes] section). Currently empty - I'm unsure whether this will be useful or not. (This used to be commit 15805164fe77b127372eba1ec51c70758467adee)
1999-04-04 10:22:22 +04:00
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
/* Allow %S to be used by force user. */
pstring_sub(fuser,"%S",service);
Call VFS initialisation, share connect and share disconnect functions. Initialise VFS function pointers with symbols from the 'vfs object' file, use disk pointers otherwise. Added a hook to check for a 'default' VFS share (like the [homes] section). Currently empty - I'm unsure whether this will be useful or not. (This used to be commit 15805164fe77b127372eba1ec51c70758467adee)
1999-04-04 10:22:22 +04:00
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
pass2 = (struct passwd *)Get_Pwnam(fuser,True);
if (pass2) {
conn->uid = pass2->pw_uid;
conn->gid = pass2->pw_gid;
string_set(&conn->user,fuser);
fstrcpy(user,fuser);
conn->force_user = True;
DEBUG(3,("Forced user %s\n",fuser));
} else {
DEBUG(1,("Couldn't find user %s\n",fuser));
}
Call VFS initialisation, share connect and share disconnect functions. Initialise VFS function pointers with symbols from the 'vfs object' file, use disk pointers otherwise. Added a hook to check for a 'default' VFS share (like the [homes] section). Currently empty - I'm unsure whether this will be useful or not. (This used to be commit 15805164fe77b127372eba1ec51c70758467adee)
1999-04-04 10:22:22 +04:00
}
fixed "admin users" option with new security code (This used to be commit b69c5de6bad9fad3aed1280e7d12fbfed276a16f)
2000-08-31 11:11:45 +04:00
/* admin users always run as uid=0 */
if (conn->admin_user) {
conn->uid = 0;
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
#ifdef HAVE_GETGRNAM
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
/*
* If force group is true, then override
* any groupid stored for the connecting user.
*/
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
if (*lp_force_group(snum)) {
Fix to allow smbd to call winbindd if it is running for all group enumeration, falling back to the UNIX calls on error. This should fix all problems with smbd enumerating all users in all groups in all trusted domains via winbindd. Also changed GETDC to query 1C name rather than 1b name as only the PDC registers 1b. Jeremy. (This used to be commit 5b0038a2afd8abbd6fd4a58f5477a40d1926d498)
2000-10-13 05:59:14 +04:00
gid_t gid;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
pstring gname;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
pstring tmp_gname;
BOOL user_must_be_member = False;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
StrnCpy(tmp_gname,lp_force_group(snum),sizeof(pstring)-1);
if (tmp_gname[0] == '+') {
user_must_be_member = True;
StrnCpy(gname,&tmp_gname[1],sizeof(pstring)-2);
} else {
StrnCpy(gname,tmp_gname,sizeof(pstring)-1);
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/* default service may be a group name */
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
pstring_sub(gname,"%S",service);
Fix to allow smbd to call winbindd if it is running for all group enumeration, falling back to the UNIX calls on error. This should fix all problems with smbd enumerating all users in all groups in all trusted domains via winbindd. Also changed GETDC to query 1C name rather than 1b name as only the PDC registers 1b. Jeremy. (This used to be commit 5b0038a2afd8abbd6fd4a58f5477a40d1926d498)
2000-10-13 05:59:14 +04:00
gid = nametogid(gname);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
Fix to allow smbd to call winbindd if it is running for all group enumeration, falling back to the UNIX calls on error. This should fix all problems with smbd enumerating all users in all groups in all trusted domains via winbindd. Also changed GETDC to query 1C name rather than 1b name as only the PDC registers 1b. Jeremy. (This used to be commit 5b0038a2afd8abbd6fd4a58f5477a40d1926d498)
2000-10-13 05:59:14 +04:00
if (gid != (gid_t)-1) {
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
/*
* If the user has been forced and the forced group starts
* with a '+', then we only set the group to be the forced
* group if the forced user is a member of that group.
* Otherwise, the meaning of the '+' would be ignored.
*/
if (conn->force_user && user_must_be_member) {
Fix to allow smbd to call winbindd if it is running for all group enumeration, falling back to the UNIX calls on error. This should fix all problems with smbd enumerating all users in all groups in all trusted domains via winbindd. Also changed GETDC to query 1C name rather than 1b name as only the PDC registers 1b. Jeremy. (This used to be commit 5b0038a2afd8abbd6fd4a58f5477a40d1926d498)
2000-10-13 05:59:14 +04:00
if (user_in_group_list( user, gname )) {
conn->gid = gid;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
DEBUG(3,("Forced group %s for member %s\n",gname,user));
}
} else {
Fix to allow smbd to call winbindd if it is running for all group enumeration, falling back to the UNIX calls on error. This should fix all problems with smbd enumerating all users in all groups in all trusted domains via winbindd. Also changed GETDC to query 1C name rather than 1b name as only the PDC registers 1b. Jeremy. (This used to be commit 5b0038a2afd8abbd6fd4a58f5477a40d1926d498)
2000-10-13 05:59:14 +04:00
conn->gid = gid;
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
DEBUG(3,("Forced group %s\n",gname));
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
} else {
DEBUG(1,("Couldn't find group %s\n",gname));
}
}
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
#endif /* HAVE_GETGRNAM */
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
{
pstring s;
pstrcpy(s,lp_pathname(snum));
- got rid of the "passive" option - cleaned up the standard_sub_*() calls a lot (This used to be commit 2c2d95d77d3667eaa9252506a82b9054b0d0e01c)
2000-04-19 08:01:16 +04:00
standard_sub_conn(conn,s);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
string_set(&conn->connectpath,s);
DEBUG(3,("Connect path is %s\n",s));
}
/* groups stuff added by ih */
conn->ngroups = 0;
conn->groups = NULL;
fixed the problem with messages not getting through the problem had nothing to do with being your own pid, it was instead a problem with IPC$ connections not being registered in the connections database and an incorrect test for -1 in the messaging code. These changes also mean that IPC$ shares now show up in smbstatus. That is probably a good thing. (This used to be commit 3575ad10985a18f897e38179ca69fa9a49a7ea02)
2000-11-17 06:15:18 +03:00
/* Find all the groups this uid is in and
store them. Used by become_user() */
initialise_groups(conn->user, conn->uid, conn->gid);
get_current_groups(&conn->ngroups,&conn->groups);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
fixed the problem with messages not getting through the problem had nothing to do with being your own pid, it was instead a problem with IPC$ connections not being registered in the connections database and an incorrect test for -1 in the messaging code. These changes also mean that IPC$ shares now show up in smbstatus. That is probably a good thing. (This used to be commit 3575ad10985a18f897e38179ca69fa9a49a7ea02)
2000-11-17 06:15:18 +03:00
/* check number of connections */
if (!claim_connection(conn,
lp_servicename(SNUM(conn)),
lp_max_connections(SNUM(conn)),
False)) {
DEBUG(1,("too many connections - rejected\n"));
*ecode = ERRnoresource;
conn_free(conn);
return NULL;
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
Changes from APPLIANCE_HEAD: source/Makefile.in - changes to ctags and etags rules that somehow got lost along the way. source/include/proto.h - make proto source/smbd/sec_ctx.c source/smbd/password.c - merge debugs for debugging user groups and NT token stuff. source/lib/util_str.c - capitalise domain name returned from parse_domain_user() source/nsswitch/wb_client.c - fix broken conditional in debug statement. source/include/rpc_secdes.h source/include/rpc_spoolss.h source/printing/nt_printing.c source/lib/util_seaccess.c - fix printer permission bugs related to ACE masks for printers. This adds mapping of generic access rights to object specific rights for NT printers. Still need to work out whether or not to ignore ACEs with certain flags set, though. See comments in util_seaccess.c:check_ace() for details. source/printing/nt_printing.c source/printing/printing.c - use PRINTER_ACCESS_ADMINISTER instead of JOB_ACCESS_ADMINISTER until we sort out printer/printjob permission stuff. (This used to be commit 1dba9c5cd1e6389734c648f6903abcb7c8d5b2f0)
2001-01-04 22:27:08 +03:00
conn->nt_user_token = create_nt_token(conn->uid, conn->gid,
Compile fix for new arg to create_nt_token() (This used to be commit 806185ca8cc8d28f16745a1db9427f52eb8d22e4)
2000-12-12 08:29:47 +03:00
conn->ngroups, conn->groups,
guest);
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
Fix for core dump in security = share code with new share security db. Jeremy. (This used to be commit 20b13bafdff2fd7be9219ed164e7fe91b597298d)
2001-04-12 01:19:25 +04:00
/*
* New code to check if there's a share security descripter
* added from NT server manager. This is done after the
* smb.conf checks are done as we need a uid and token. JRA.
*/
{
BOOL can_write = share_access_check(conn, snum, vuid, FILE_WRITE_DATA);
if (!can_write) {
if (!share_access_check(conn, snum, vuid, FILE_READ_DATA)) {
/* No access, read or write. */
*ecode = ERRaccess;
DEBUG(0,( "make_connection: connection to %s denied due to security descriptor.\n",
service ));
conn_free(conn);
return NULL;
} else {
conn->read_only = True;
}
}
}
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
/* Initialise VFS function pointers */
if (*lp_vfsobj(SNUM(conn))) {
#ifdef HAVE_LIBDL
/* Loadable object file */
if (!vfs_init_custom(conn)) {
Added a VFS version return to init call. Allows smbd to fail an init if versions don't match. Jeremy. (This used to be commit d0fbb4f5d999abade8930cc6fff231a2af6cccfb)
2000-11-07 00:44:33 +03:00
DEBUG(0, ("vfs_init failed\n"));
conn_free(conn);
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
return NULL;
}
#else
DEBUG(0, ("No libdl present - cannot use VFS objects\n"));
conn_free(conn);
return NULL;
#endif
} else {
/* Normal share - initialise with disk access functions */
vfs_init_default(conn);
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
/* execute any "root preexec = " line */
if (*lp_rootpreexec(SNUM(conn))) {
pstring cmd;
pstrcpy(cmd,lp_rootpreexec(SNUM(conn)));
- got rid of the "passive" option - cleaned up the standard_sub_*() calls a lot (This used to be commit 2c2d95d77d3667eaa9252506a82b9054b0d0e01c)
2000-04-19 08:01:16 +04:00
standard_sub_conn(conn,cmd);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
DEBUG(5,("cmd=%s\n",cmd));
Merge of Andrew's changes in 2.2. Jeremy. (This used to be commit fc76681812b1469208ad6c8847afdfc68bc6db49)
2001-04-13 23:12:06 +04:00
ret = smbrun(cmd,NULL);
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
if (ret != 0 && lp_rootpreexec_close(SNUM(conn))) {
DEBUG(1,("preexec gave %d - failing connection\n", ret));
conn_free(conn);
*ecode = ERRsrverror;
return NULL;
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
if (!become_user(conn, conn->vuid)) {
DEBUG(0,("Can't become connected user!\n"));
fixed the problem with messages not getting through the problem had nothing to do with being your own pid, it was instead a problem with IPC$ connections not being registered in the connections database and an incorrect test for -1 in the messaging code. These changes also mean that IPC$ shares now show up in smbstatus. That is probably a good thing. (This used to be commit 3575ad10985a18f897e38179ca69fa9a49a7ea02)
2000-11-17 06:15:18 +03:00
yield_connection(conn,
lp_servicename(SNUM(conn)),
lp_max_connections(SNUM(conn)));
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
conn_free(conn);
*ecode = ERRbadpw;
return NULL;
}
Restructuring of the code to remove dos_ChDir/dos_GetWd and re-vector them through the VFS. All file access/directory access code in smbd should now go via the vfs. Added vfs_chown/vfs_chmod calls. Still looking at vfs_get_nt_acl() vfs_set_nt_acl() call API design. Jeremy. (This used to be commit f96625ec124adb6e110dc54632e006b3620a962b)
2000-09-27 23:09:59 +04:00
if (vfs_ChDir(conn,conn->connectpath) != 0) {
Patch from Massimo Sivilotti <mass@tanner.com> to log remote machine/ip on connection fail. Jeremy. (This used to be commit 07cee46d1de1caaf6f9f1b6139dd21bcc5d67e8e)
2001-03-27 03:21:29 +04:00
DEBUG(0,("%s (%s) Can't change directory to %s (%s)\n",
remote_machine, conn->client_address,
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
conn->connectpath,strerror(errno)));
unbecome_user();
fixed the problem with messages not getting through the problem had nothing to do with being your own pid, it was instead a problem with IPC$ connections not being registered in the connections database and an incorrect test for -1 in the messaging code. These changes also mean that IPC$ shares now show up in smbstatus. That is probably a good thing. (This used to be commit 3575ad10985a18f897e38179ca69fa9a49a7ea02)
2000-11-17 06:15:18 +03:00
yield_connection(conn,
lp_servicename(SNUM(conn)),
lp_max_connections(SNUM(conn)));
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
conn_free(conn);
Fixed bug where file access was allowed on IPC$ share. Return correct error codes on invalid share name. Jeremy. (This used to be commit 420d6bc4809cef9d74354175d0fa956ab4e8ac3c)
2000-05-23 05:27:19 +04:00
*ecode = ERRnosuchshare;
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
return NULL;
}
string_set(&conn->origpath,conn->connectpath);
#if SOFTLINK_OPTIMISATION
/* resolve any soft links early */
{
pstring s;
pstrcpy(s,conn->connectpath);
Restructuring of the code to remove dos_ChDir/dos_GetWd and re-vector them through the VFS. All file access/directory access code in smbd should now go via the vfs. Added vfs_chown/vfs_chmod calls. Still looking at vfs_get_nt_acl() vfs_set_nt_acl() call API design. Jeremy. (This used to be commit f96625ec124adb6e110dc54632e006b3620a962b)
2000-09-27 23:09:59 +04:00
vfs_GetWd(conn,s);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
string_set(&conn->connectpath,s);
Restructuring of the code to remove dos_ChDir/dos_GetWd and re-vector them through the VFS. All file access/directory access code in smbd should now go via the vfs. Added vfs_chown/vfs_chmod calls. Still looking at vfs_get_nt_acl() vfs_set_nt_acl() call API design. Jeremy. (This used to be commit f96625ec124adb6e110dc54632e006b3620a962b)
2000-09-27 23:09:59 +04:00
vfs_ChDir(conn,conn->connectpath);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
#endif
add_session_user(user);
/* execute any "preexec = " line */
if (*lp_preexec(SNUM(conn))) {
pstring cmd;
pstrcpy(cmd,lp_preexec(SNUM(conn)));
- got rid of the "passive" option - cleaned up the standard_sub_*() calls a lot (This used to be commit 2c2d95d77d3667eaa9252506a82b9054b0d0e01c)
2000-04-19 08:01:16 +04:00
standard_sub_conn(conn,cmd);
Merge of Andrew's changes in 2.2. Jeremy. (This used to be commit fc76681812b1469208ad6c8847afdfc68bc6db49)
2001-04-13 23:12:06 +04:00
ret = smbrun(cmd,NULL);
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
if (ret != 0 && lp_preexec_close(SNUM(conn))) {
DEBUG(1,("preexec gave %d - failing connection\n", ret));
conn_free(conn);
*ecode = ERRsrverror;
return NULL;
}
}
/*
* Print out the 'connected as' stuff here as we need
* to know the effective uid and gid we will be using.
*/
if( DEBUGLVL( IS_IPC(conn) ? 3 : 1 ) ) {
dbgtext( "%s (%s) ", remote_machine, conn->client_address );
dbgtext( "connect to service %s ", lp_servicename(SNUM(conn)) );
dbgtext( "as user %s ", user );
dbgtext( "(uid=%d, gid=%d) ", (int)geteuid(), (int)getegid() );
Added sys_fork() and sys_getpid() functions to stop the overhead of doing a system call every time we want to just get our pid. Jeremy. (This used to be commit 148628b616b5c29ba6340d65fc3ddbcabba6e67a)
2000-05-02 06:23:41 +04:00
dbgtext( "(pid %d)\n", (int)sys_getpid() );
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
/* we've finished with the sensitive stuff */
unbecome_user();
/* Add veto/hide lists */
if (!IS_IPC(conn) && !IS_PRINT(conn)) {
set_namearray( &conn->veto_list, lp_veto_files(SNUM(conn)));
set_namearray( &conn->hide_list, lp_hide_files(SNUM(conn)));
set_namearray( &conn->veto_oplock_list, lp_veto_oplocks(SNUM(conn)));
}
Mega-VFS merge. Yeah baby! Synopsis: change every disk access function to work through a vfs_ops structure contained in the connection_struct. (This used to be commit 3aad500c0fb61232ed3431ff4b743b5d18ec852f)
2000-02-03 08:17:25 +03:00
/* Invoke VFS make connection hook */
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
if (conn->vfs_ops.connect) {
Restructuring of vfs layer to include a "this" pointer - can be an fsp or a conn struct depending on the call. We need this to have a clean NT ACL call interface. This will break any existing VFS libraries (that's why this is pre-release code). Andrew gets credit for this one :-) :-). In addition - added Herb's WITH_PROFILE changes - Herb - please examine the changes I've made to the smbd/reply.c code you added. The original code was very ugly and I have replaced it with a START_PROFILE(x)/END_PROFILE(x) pair using the preprocessor. Please check this compiles ok with the --with-profile switch. Jeremy. (This used to be commit b07611f8159b0b3f42e7e02611be9f4d56de96f5)
2000-10-06 07:21:49 +04:00
if (conn->vfs_ops.connect(conn, service, user) < 0)
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
return NULL;
}
Mega-VFS merge. Yeah baby! Synopsis: change every disk access function to work through a vfs_ops structure contained in the connection_struct. (This used to be commit 3aad500c0fb61232ed3431ff4b743b5d18ec852f)
2000-02-03 08:17:25 +03:00
Added an NT_USER_TOKEN structure that is copied/passed around associated with the current user. This will allow se_access_check() to quickly do a SD check without having to translate uid/gid's to SIDs. Still needs work on pipe calls. Jeremy. (This used to be commit e28d01b744b3dbd33e0e54af4e7f426fa8c082b8)
2000-08-04 02:38:43 +04:00
return(conn);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
/****************************************************************************
close a cnum
****************************************************************************/
void close_cnum(connection_struct *conn, uint16 vuid)
{
DirCacheFlush(SNUM(conn));
unbecome_user();
DEBUG(IS_IPC(conn)?3:1, ("%s (%s) closed connection to service %s\n",
first pass at updating head branch to be to be the same as the SAMBA_2_0 branch (This used to be commit 453a822a76780063dff23526c35408866d0c0154)
1999-12-13 16:27:58 +03:00
remote_machine,conn->client_address,
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
lp_servicename(SNUM(conn))));
Mega-VFS merge. Yeah baby! Synopsis: change every disk access function to work through a vfs_ops structure contained in the connection_struct. (This used to be commit 3aad500c0fb61232ed3431ff4b743b5d18ec852f)
2000-02-03 08:17:25 +03:00
if (conn->vfs_ops.disconnect != NULL) {
/* Call VFS disconnect hook */
Restructuring of vfs layer to include a "this" pointer - can be an fsp or a conn struct depending on the call. We need this to have a clean NT ACL call interface. This will break any existing VFS libraries (that's why this is pre-release code). Andrew gets credit for this one :-) :-). In addition - added Herb's WITH_PROFILE changes - Herb - please examine the changes I've made to the smbd/reply.c code you added. The original code was very ugly and I have replaced it with a START_PROFILE(x)/END_PROFILE(x) pair using the preprocessor. Please check this compiles ok with the --with-profile switch. Jeremy. (This used to be commit b07611f8159b0b3f42e7e02611be9f4d56de96f5)
2000-10-06 07:21:49 +04:00
conn->vfs_ops.disconnect(conn);
Mega-VFS merge. Yeah baby! Synopsis: change every disk access function to work through a vfs_ops structure contained in the connection_struct. (This used to be commit 3aad500c0fb61232ed3431ff4b743b5d18ec852f)
2000-02-03 08:17:25 +03:00
}
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
yield_connection(conn,
lp_servicename(SNUM(conn)),
lp_max_connections(SNUM(conn)));
file_close_conn(conn);
dptr_closecnum(conn);
/* execute any "postexec = " line */
if (*lp_postexec(SNUM(conn)) &&
become_user(conn, vuid)) {
pstring cmd;
pstrcpy(cmd,lp_postexec(SNUM(conn)));
- got rid of the "passive" option - cleaned up the standard_sub_*() calls a lot (This used to be commit 2c2d95d77d3667eaa9252506a82b9054b0d0e01c)
2000-04-19 08:01:16 +04:00
standard_sub_conn(conn,cmd);
Merge of Andrew's changes in 2.2. Jeremy. (This used to be commit fc76681812b1469208ad6c8847afdfc68bc6db49)
2001-04-13 23:12:06 +04:00
smbrun(cmd,NULL);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
unbecome_user();
}
unbecome_user();
/* execute any "root postexec = " line */
if (*lp_rootpostexec(SNUM(conn))) {
pstring cmd;
pstrcpy(cmd,lp_rootpostexec(SNUM(conn)));
- got rid of the "passive" option - cleaned up the standard_sub_*() calls a lot (This used to be commit 2c2d95d77d3667eaa9252506a82b9054b0d0e01c)
2000-04-19 08:01:16 +04:00
standard_sub_conn(conn,cmd);
Merge of Andrew's changes in 2.2. Jeremy. (This used to be commit fc76681812b1469208ad6c8847afdfc68bc6db49)
2001-04-13 23:12:06 +04:00
smbrun(cmd,NULL);
this completes the splitup of server.c. the splitup was done with an axe, not a scalpel, so there are some rough edges. I mostly wanted to get the general form right with fine tuning of what goes where to come later. Still, this is better than what we had before where server.c was a general repository for anything that didn't fit elsewhere. (This used to be commit a6d194886a4a5f7507fa37289ff96c1be56f14a6)
1998-08-17 17:11:34 +04:00
}
conn_free(conn);
}
added a close-share smbcontrol message that forcibly closes a share in smbd (to allow unmount) (This used to be commit 15b17a80db605a55f667c95fb7e316877a441887)
2001-06-20 07:05:09 +04:00
Copy Permalink