Merge from appliance head of JR's changes for driver versioning.

Jeremy.
(This used to be commit cdbd2e99775642dc2e92004be9014bf38a92d80f)

4 files changed, 141 insertions, 73 deletions

diff --git a/source3/smbd/reply.c b/source3/smbd/reply.c
index fa8aa11277..0b3b5bbe27 100644
--- a/source3/smbd/reply.c
+++ b/source3/smbd/reply.c
@@ -1882,14 +1882,13 @@ static BOOL can_delete(char *fname,connection_struct *conn, int dirtype)
 }
 
 /****************************************************************************
- Reply to a unlink
+ The guts of the unlink command, split out so it may be called by the NT SMB
+ code.
 ****************************************************************************/
 
-int reply_unlink(connection_struct *conn, char *inbuf,char *outbuf, int dum_size, int dum_buffsize)
+int unlink_internals(connection_struct *conn, char *inbuf,char *outbuf,
+					 int dirtype, char *name)
 {
-  int outsize = 0;
-  pstring name;
-  int dirtype;
   pstring directory;
   pstring mask;
   char *p;
@@ -1900,18 +1899,9 @@ int reply_unlink(connection_struct *conn, char *inbuf,char *outbuf, int dum_size
   BOOL bad_path = False;
   BOOL rc = True;
   SMB_STRUCT_STAT sbuf;
-  START_PROFILE(SMBunlink);
 
   *directory = *mask = 0;
 
-  dirtype = SVAL(inbuf,smb_vwv0);
-  
-  pstrcpy(name,smb_buf(inbuf) + 1);
-   
-  RESOLVE_DFSPATH(name, conn, inbuf, outbuf);
-
-  DEBUG(3,("reply_unlink : %s\n",name));
-   
   rc = unix_convert(name,conn,0,&bad_path,&sbuf);
 
   p = strrchr(name,'/');
@@ -1975,29 +1965,58 @@ int reply_unlink(connection_struct *conn, char *inbuf,char *outbuf, int dum_size
 	    slprintf(fname,sizeof(fname)-1, "%s/%s",directory,dname);
 	    if (!can_delete(fname,conn,dirtype)) continue;
 	    if (!vfs_unlink(conn,fname)) count++;
-	    DEBUG(3,("reply_unlink : doing unlink on %s\n",fname));
+	    DEBUG(3,("unlink_internals: succesful unlink [%s]\n",fname));
 	  }
 	CloseDir(dirptr);
       }
   }
   
   if (count == 0) {
-    if (exists) {
-      END_PROFILE(SMBunlink);
+    if (exists)
       return(ERROR(ERRDOS,error));
-    } else
-    {
-      if((errno == ENOENT) && bad_path)
-      {
+    else {
+      if((errno == ENOENT) && bad_path) {
         unix_ERR_class = ERRDOS;
         unix_ERR_code = ERRbadpath;
       }
-      END_PROFILE(SMBunlink);
       return(UNIXERROR(ERRDOS,error));
     }
   }
   
+  return 0;
+}
+
+/****************************************************************************
+ Reply to a unlink
+****************************************************************************/
+
+int reply_unlink(connection_struct *conn, char *inbuf,char *outbuf, int dum_size, int dum_buffsize)
+{
+  int outsize = 0;
+  pstring name;
+  int dirtype;
+  START_PROFILE(SMBunlink);
+
+  dirtype = SVAL(inbuf,smb_vwv0);
+
+  pstrcpy(name,smb_buf(inbuf) + 1);
+
+  RESOLVE_DFSPATH(name, conn, inbuf, outbuf);
+
+  DEBUG(3,("reply_unlink : %s\n",name));
+
+  outsize = unlink_internals(conn, inbuf, outbuf, dirtype, name);
+  if(outsize == 0) {
+
+    /*
+     * Win2k needs a changenotify request response before it will
+     * update after a rename..
+     */
+
+    process_pending_change_notify_queue((time_t)0);
+
   outsize = set_message(outbuf,0,0,True);
+  }
   
   END_PROFILE(SMBunlink);
   return(outsize);
@@ -3589,7 +3608,6 @@ static BOOL can_rename(char *fname,connection_struct *conn)
 
   if (conn->vfs_ops.lstat(conn,dos_to_unix(fname,False),&sbuf) != 0) return(False);
   if (!check_file_sharing(conn,fname,True)) return(False);
-
   return(True);
 }
 
@@ -3718,6 +3736,7 @@ int rename_internals(connection_struct *conn,
 			 * file with the same name so don't check for
 			 * vfs_file_exist().
 			 */
+
 			if(resolve_wildcards(directory,newname) &&
 			   can_rename(directory,conn) &&
 			   !conn->vfs_ops.rename(conn,zdirectory,
diff --git a/source3/smbd/sec_ctx.c b/source3/smbd/sec_ctx.c
index ef446ed759..f3adcfce8f 100644
--- a/source3/smbd/sec_ctx.c
+++ b/source3/smbd/sec_ctx.c
@@ -221,7 +221,7 @@ BOOL initialise_groups(char *user, uid_t uid, gid_t gid)
 
 	/* Call initgroups() to get user groups */
 
-	if (initgroups(user,gid) == -1) {
+	if (winbind_initgroups(user,gid) == -1) {
 		DEBUG(0,("Unable to initgroups. Error was %s\n", strerror(errno) ));
 		if (getuid() == 0) {
 			if (gid < 0 || gid > 32767 || uid < 0 || uid > 32767) {
diff --git a/source3/smbd/uid.c b/source3/smbd/uid.c
index d82edcbfae..008765cde1 100644
--- a/source3/smbd/uid.c
+++ b/source3/smbd/uid.c
@@ -243,8 +243,8 @@ BOOL unbecome_authenticated_pipe_user(pipes_struct *p)
 	return pop_sec_ctx();
 }
 
-
 /* Temporarily become a root user.  Must match with unbecome_root(). */
+
 void become_root(void)
 {
 	push_sec_ctx();
@@ -267,25 +267,49 @@ BOOL lookup_name(char *name, DOM_SID *psid, enum SID_NAME_USE *name_type)
 {
 	extern pstring global_myname;
 	fstring sid;
+	char *sep = lp_winbind_separator();
 
 	if (!winbind_lookup_name(name, psid, name_type)) {
 		BOOL ret;
 
-		DEBUG(10,("lookup_name: winbind lookup for %s failed - trying local\n", name ));
+		DEBUG(10, ("lookup_name: winbind lookup for %s failed - trying local\n", name));
+
+		/* If we are looking up a domain user, make sure it is
+		   for the local machine only */
+
+		if (strchr(name, sep[0]) || strchr(name, '\\')) {
+			fstring domain, username;
+
+			split_domain_name(name, domain, username);
+
+			if (strcasecmp(global_myname, domain) != 0) {
+				DEBUG(5, ("domain %s is not local\n", domain));
+				return False;
+			}
+
+			ret = local_lookup_name(domain, username, psid, 
+						name_type);
+		} else {
+
+			ret = local_lookup_name(global_myname, name, psid, 
+						name_type);
+		}
+
+		if (ret) {
+			DEBUG(10,
+			      ("lookup_name: (local) %s -> SID %s (type %u)\n",
+			       name, sid_to_string(sid,psid),
+			       (unsigned int)*name_type ));
+		} else {
+			DEBUG(10,("lookup name: (local) %s failed.\n", name));
+		}
 
-		ret = local_lookup_name(global_myname, name, psid, name_type);
-		if (ret)
-			DEBUG(10,("lookup_name : (local) %s -> SID %s (type %u)\n",
-				name, sid_to_string(sid,psid),
-				(unsigned int)*name_type ));
-		else
-			DEBUG(10,("lookup name : (local) %s failed.\n",
-					name ));
 		return ret;
 	}
 
-	DEBUG(10,("lookup_name (winbindd): %s -> SID %s (type %u)\n",
-		name, sid_to_string(sid,psid), (unsigned int)*name_type ));
+		DEBUG(10,("lookup_name (winbindd): %s -> SID %s (type %u)\n",
+			  name, sid_to_string(sid, psid), 
+			  (unsigned int)*name_type));
 	return True;
 }
 
@@ -300,7 +324,7 @@ BOOL lookup_sid(DOM_SID *sid, fstring dom_name, fstring name, enum SID_NAME_USE
 		return False;
 
 	/* Check if this is our own sid.  This should perhaps be done by
-		winbind?  For the moment handle it here. */
+	   winbind?  For the moment handle it here. */
 
 	if (sid->num_auths == 5) {
 		DOM_SID tmp_sid;
diff --git a/source3/smbd/vfs.c b/source3/smbd/vfs.c
index 99c8e26fa8..77b42361ed 100644
--- a/source3/smbd/vfs.c
+++ b/source3/smbd/vfs.c
@@ -1,19 +1,19 @@
-/* 
+/*
    Unix SMB/Netbios implementation.
    Version 1.9.
    VFS initialisation and support functions
    Copyright (C) Tim Potter 1999
-   
+
    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.
@@ -36,7 +36,7 @@ struct vfs_syminfo {
 
 struct vfs_ops default_vfs_ops = {
 
-	/* Disk operations */        
+	/* Disk operations */
 
 	vfswrap_dummy_connect,
 	vfswrap_dummy_disconnect,
@@ -112,7 +112,7 @@ BOOL vfs_init_custom(connection_struct *conn)
     init_fptr = (struct vfs_ops *(*)(int *))dlsym(conn->dl_handle, "vfs_init");
 
     if (init_fptr == NULL) {
-		DEBUG(0, ("No vfs_init() symbol found in %s\n", 
+		DEBUG(0, ("No vfs_init() symbol found in %s\n",
 		  lp_vfsobj(SNUM(conn))));
 		return False;
     }
@@ -132,10 +132,10 @@ BOOL vfs_init_custom(connection_struct *conn)
 
     /* Fill in unused operations with default (disk based) ones.
        There's probably a neater way to do this then a whole bunch of
-       if statements. */ 
+       if statements. */
 
     memcpy(&conn->vfs_ops, ops, sizeof(struct vfs_ops));
-    
+
     if (conn->vfs_ops.connect == NULL) {
 	conn->vfs_ops.connect = default_vfs_ops.connect;
     }
@@ -179,47 +179,47 @@ BOOL vfs_init_custom(connection_struct *conn)
     if (conn->vfs_ops.read == NULL) {
 	conn->vfs_ops.read = default_vfs_ops.read;
     }
-    
+
     if (conn->vfs_ops.write == NULL) {
 	conn->vfs_ops.write = default_vfs_ops.write;
     }
-    
+
     if (conn->vfs_ops.lseek == NULL) {
 	conn->vfs_ops.lseek = default_vfs_ops.lseek;
     }
-    
+
     if (conn->vfs_ops.rename == NULL) {
 	conn->vfs_ops.rename = default_vfs_ops.rename;
     }
-    
+
     if (conn->vfs_ops.fsync == NULL) {
 	conn->vfs_ops.fsync = default_vfs_ops.fsync;
     }
-    
+
     if (conn->vfs_ops.stat == NULL) {
 	conn->vfs_ops.stat = default_vfs_ops.stat;
     }
-    
+
     if (conn->vfs_ops.fstat == NULL) {
 	conn->vfs_ops.fstat = default_vfs_ops.fstat;
     }
-    
+
     if (conn->vfs_ops.lstat == NULL) {
 	conn->vfs_ops.lstat = default_vfs_ops.lstat;
     }
-    
+
     if (conn->vfs_ops.unlink == NULL) {
 	conn->vfs_ops.unlink = default_vfs_ops.unlink;
     }
-    
+
     if (conn->vfs_ops.chmod == NULL) {
 	conn->vfs_ops.chmod = default_vfs_ops.chmod;
     }
-    
+
     if (conn->vfs_ops.chown == NULL) {
 	conn->vfs_ops.chown = default_vfs_ops.chown;
     }
-    
+
     if (conn->vfs_ops.chdir == NULL) {
 	conn->vfs_ops.chdir = default_vfs_ops.chdir;
     }
@@ -227,15 +227,15 @@ BOOL vfs_init_custom(connection_struct *conn)
     if (conn->vfs_ops.getwd == NULL) {
 	conn->vfs_ops.getwd = default_vfs_ops.getwd;
     }
-    
+
     if (conn->vfs_ops.utime == NULL) {
 	conn->vfs_ops.utime = default_vfs_ops.utime;
     }
-    
+
     if (conn->vfs_ops.ftruncate == NULL) {
 	conn->vfs_ops.ftruncate = default_vfs_ops.ftruncate;
     }
-    
+
     if (conn->vfs_ops.lock == NULL) {
 	conn->vfs_ops.lock = default_vfs_ops.lock;
     }
@@ -255,7 +255,7 @@ BOOL vfs_init_custom(connection_struct *conn)
     if (conn->vfs_ops.set_nt_acl == NULL) {
 	conn->vfs_ops.set_nt_acl = default_vfs_ops.set_nt_acl;
     }
-    
+
     return True;
 }
 #endif
@@ -272,7 +272,7 @@ BOOL vfs_directory_exist(connection_struct *conn, char *dname, SMB_STRUCT_STAT *
 	if (!st)
 		st = &st2;
 
-	if (vfs_stat(conn,dname,st) != 0) 
+	if (vfs_stat(conn,dname,st) != 0)
 		return(False);
 
 	ret = S_ISDIR(st->st_mode);
@@ -294,7 +294,7 @@ int vfs_mkdir(connection_struct *conn, char *fname, mode_t mode)
 
 	pstrcpy(name,dos_to_unix(fname,False)); /* paranoia copy */
 	if(!(ret=conn->vfs_ops.mkdir(conn,name,mode))) {
-		/* 
+		/*
 		 * Check if high bits should have been set,
 		 * then (if bits are missing): add them.
 		 * Consider bits automagically set by UNIX, i.e. SGID bit from parent dir.
@@ -331,14 +331,39 @@ BOOL vfs_file_exist(connection_struct *conn,char *fname,SMB_STRUCT_STAT *sbuf)
 		sbuf = &st;
 
 	ZERO_STRUCTP(sbuf);
- 
-	if (vfs_stat(conn,fname,sbuf) != 0) 
+
+	if (vfs_stat(conn,fname,sbuf) != 0)
 		return(False);
 
 	return(S_ISREG(sbuf->st_mode));
 }
 
 /****************************************************************************
+ Read data from fsp on the vfs. (note: EINTR re-read differs from vfs_write_data)
+****************************************************************************/
+
+ssize_t vfs_read_data(files_struct *fsp, char *buf, size_t byte_count)
+{
+	size_t total=0;
+
+	while (total < byte_count)
+	{
+		ssize_t ret = fsp->conn->vfs_ops.read(fsp, fsp->fd, buf + total,
+											  byte_count - total);
+
+		if (ret == 0) return total;
+		if (ret == -1) {
+			if (errno == EINTR)
+				continue;
+			else
+				return -1;
+		}
+		total += ret;
+	}
+	return (ssize_t)total;
+}
+
+/****************************************************************************
  Write data to a fd on the vfs.
 ****************************************************************************/
 
@@ -363,11 +388,11 @@ ssize_t vfs_write_data(files_struct *fsp,char *buffer,size_t N)
  Transfer some data between two file_struct's.
 ****************************************************************************/
 
-SMB_OFF_T vfs_transfer_file(int in_fd, files_struct *in_fsp, 
+SMB_OFF_T vfs_transfer_file(int in_fd, files_struct *in_fsp,
 			    int out_fd, files_struct *out_fsp,
 			    SMB_OFF_T n, char *header, int headlen, int align)
 {
-  static char *buf=NULL;  
+  static char *buf=NULL;
   static int size=0;
   char *buf1,*abuf;
   SMB_OFF_T total = 0;
@@ -425,7 +450,7 @@ SMB_OFF_T vfs_transfer_file(int in_fd, files_struct *in_fsp,
     }
 
     if (s > ret) {
-      ret += in_fsp ? 
+      ret += in_fsp ?
 	  in_fsp->conn->vfs_ops.read(in_fsp,in_fsp->fd,buf1+ret,s-ret) : read(in_fd,buf1+ret,s-ret);
     }
 
@@ -459,7 +484,7 @@ char *vfs_readdirname(connection_struct *conn, void *p)
 
 	if (!p)
 		return(NULL);
-  
+
 	ptr = (struct dirent *)conn->vfs_ops.readdir(conn,p);
 	if (!ptr)
 		return(NULL);
@@ -497,7 +522,7 @@ static BOOL handle_vfs_option(char *pszParmValue, char **ptr)
 {
     struct vfs_options *new_option, **options = (struct vfs_options **)ptr;
     int i;
-    
+
     /* Create new vfs option */
 
     new_option = (struct vfs_options *)malloc(sizeof(*new_option));
@@ -508,7 +533,7 @@ static BOOL handle_vfs_option(char *pszParmValue, char **ptr)
     ZERO_STRUCTP(new_option);
 
     /* Get name and value */
-    
+
     new_option->name = strtok(pszParmValue, "=");
 
     if (new_option->name == NULL) {
@@ -714,7 +739,7 @@ char *vfs_GetWd(connection_struct *conn, char *path)
 }
 
 /*******************************************************************
- Reduce a file name, removing .. elements and checking that 
+ Reduce a file name, removing .. elements and checking that
  it is below dir in the heirachy. This uses vfs_GetWd() and so must be run
  on the system that has the referenced file system.
  Widelinks are allowed if widelinks is true.
@@ -749,7 +774,7 @@ BOOL reduce_name(connection_struct *conn, char *s,char *dir,BOOL widelinks)
 
     return(True);
   }
-  
+
   DEBUG(3,("reduce_name [%s] [%s]\n",s,dir));
 
   /* remove any double slashes */
@@ -810,7 +835,7 @@ BOOL reduce_name(connection_struct *conn, char *s,char *dir,BOOL widelinks)
   }
 
   {
-    size_t l = strlen(dir2);    
+    size_t l = strlen(dir2);
     if (dir2[l-1] == '/')
       l--;