Fix VFS layer:

1. Finally work with cascaded modules with private data storage per module
2. Convert VFS API to macro calls to simplify cascading
3. Add quota support to VFS layer (prepare to NT quota support)

Patch by Stefan (metze) Metzemacher, with review of Jelmer and me
Tested in past few weeks. Documentation to new VFS API for third-party developers to follow
(This used to be commit 91984ef5caa2d13c5d52e1f535bd3bbbae1ec978)

1 files changed, 28 insertions, 28 deletions

diff --git a/source3/smbd/reply.c b/source3/smbd/reply.c
index 9577196bfe..49a9b934b0 100644
--- a/source3/smbd/reply.c
+++ b/source3/smbd/reply.c
@@ -399,7 +399,7 @@ int reply_chkpth(connection_struct *conn, char *inbuf,char *outbuf, int dum_size
 	mode = SVAL(inbuf,smb_vwv0);
 
 	if (check_name(name,conn)) {
-		if (VALID_STAT(sbuf) || vfs_stat(conn,name,&sbuf) == 0)
+		if (VALID_STAT(sbuf) || VFS_STAT(conn,name,&sbuf) == 0)
 			if (!(ok = S_ISDIR(sbuf.st_mode)))
 				errno = ENOTDIR;
 	}
@@ -458,7 +458,7 @@ int reply_getatr(connection_struct *conn, char *inbuf,char *outbuf, int dum_size
 	} else {
 		unix_convert(fname,conn,0,&bad_path,&sbuf);
 		if (check_name(fname,conn)) {
-			if (VALID_STAT(sbuf) || vfs_stat(conn,fname,&sbuf) == 0) {
+			if (VALID_STAT(sbuf) || VFS_STAT(conn,fname,&sbuf) == 0) {
 				mode = dos_mode(conn,fname,&sbuf);
 				size = sbuf.st_size;
 				mtime = sbuf.st_mtime;
@@ -553,7 +553,7 @@ int reply_dskattr(connection_struct *conn, char *inbuf,char *outbuf, int dum_siz
 	SMB_BIG_UINT dfree,dsize,bsize;
 	START_PROFILE(SMBdskattr);
 
-	conn->vfs_ops.disk_free(conn,".",True,&bsize,&dfree,&dsize);
+	VFS_DISK_FREE(conn,".",True,&bsize,&dfree,&dsize);
   
 	outsize = set_message(outbuf,5,0,True);
 	
@@ -1128,7 +1128,7 @@ int reply_ctemp(connection_struct *conn, char *inbuf,char *outbuf, int dum_size,
 		return(UNIXERROR(ERRDOS,ERRnoaccess));
 	}
 
-	vfs_stat(conn,fname,&sbuf);
+	VFS_STAT(conn,fname,&sbuf);
 
 	/* Open file in dos compatibility share mode. */
 	/* We should fail if file does not exist. */
@@ -1227,7 +1227,7 @@ static NTSTATUS can_delete(char *fname,connection_struct *conn, int dirtype)
 	if (!CAN_WRITE(conn))
 		return NT_STATUS_MEDIA_WRITE_PROTECTED;
 
-	if (conn->vfs_ops.lstat(conn,fname,&sbuf) != 0)
+	if (VFS_LSTAT(conn,fname,&sbuf) != 0)
 		return NT_STATUS_OBJECT_NAME_NOT_FOUND;
 
 	fmode = dos_mode(conn,fname,&sbuf);
@@ -1313,7 +1313,7 @@ NTSTATUS unlink_internals(connection_struct *conn, int dirtype, char *name)
 		error = can_delete(directory,conn,dirtype);
 		if (!NT_STATUS_IS_OK(error)) return error;
 
-		if (vfs_unlink(conn,directory) == 0) {
+		if (VFS_UNLINK(conn,directory) == 0) {
 			count++;
 		}
 	} else {
@@ -1343,7 +1343,7 @@ NTSTATUS unlink_internals(connection_struct *conn, int dirtype, char *name)
 				slprintf(fname,sizeof(fname)-1, "%s/%s",directory,dname);
 				error = can_delete(fname,conn,dirtype);
 				if (!NT_STATUS_IS_OK(error)) continue;
-				if (vfs_unlink(conn,fname) == 0) count++;
+				if (VFS_UNLINK(conn,fname) == 0) count++;
 				DEBUG(3,("unlink_internals: succesful unlink [%s]\n",fname));
 			}
 			CloseDir(dirptr);
@@ -1429,7 +1429,7 @@ void send_file_readbraw(connection_struct *conn, files_struct *fsp, SMB_OFF_T st
 		header.length = 4;
 		header.free = NULL;
 
-		if ( conn->vfs_ops.sendfile( smbd_server_fd(), fsp, fsp->fd, &header, startpos, nread) == -1) {
+		if ( VFS_SENDFILE( smbd_server_fd(), fsp, fsp->fd, &header, startpos, nread) == -1) {
 			/*
 			 * Special hack for broken Linux with no 64 bit clean sendfile. If we
 			 * return ENOSYS then pretend we just got a normal read.
@@ -1554,7 +1554,7 @@ int reply_readbraw(connection_struct *conn, char *inbuf, char *outbuf, int dum_s
   
 		if (size < sizeneeded) {
 			SMB_STRUCT_STAT st;
-			if (vfs_fstat(fsp,fsp->fd,&st) == 0)
+			if (VFS_FSTAT(fsp,fsp->fd,&st) == 0)
 				size = st.st_size;
 			if (!fsp->can_write) 
 				fsp->size = size;
@@ -1723,7 +1723,7 @@ int send_file_readX(connection_struct *conn, char *inbuf,char *outbuf,int length
 		SMB_STRUCT_STAT sbuf;
 		DATA_BLOB header;
 
-		if(vfs_fstat(fsp,fsp->fd, &sbuf) == -1)
+		if(VFS_FSTAT(fsp,fsp->fd, &sbuf) == -1)
 			return(UNIXERROR(ERRDOS,ERRnoaccess));
 
 		if (startpos > sbuf.st_size)
@@ -1750,7 +1750,7 @@ int send_file_readX(connection_struct *conn, char *inbuf,char *outbuf,int length
 		header.length = data - outbuf;
 		header.free = NULL;
 
-		if ( conn->vfs_ops.sendfile( smbd_server_fd(), fsp, fsp->fd, &header, startpos, smb_maxcnt) == -1) {
+		if ( VFS_SENDFILE( smbd_server_fd(), fsp, fsp->fd, &header, startpos, smb_maxcnt) == -1) {
 			/*
 			 * Special hack for broken Linux with no 64 bit clean sendfile. If we
 			 * return ENOSYS then pretend we just got a normal read.
@@ -2262,7 +2262,7 @@ int reply_lseek(connection_struct *conn, char *inbuf,char *outbuf, int size, int
 			break;
 	}
 
-	if((res = conn->vfs_ops.lseek(fsp,fsp->fd,startpos,umode)) == -1) {
+	if((res = VFS_LSEEK(fsp,fsp->fd,startpos,umode)) == -1) {
 		/*
 		 * Check for the special case where a seek before the start
 		 * of the file sets the offset to zero. Added in the CIFS spec,
@@ -2274,7 +2274,7 @@ int reply_lseek(connection_struct *conn, char *inbuf,char *outbuf, int size, int
 
 			if(umode == SEEK_CUR) {
 
-				if((current_pos = conn->vfs_ops.lseek(fsp,fsp->fd,0,SEEK_CUR)) == -1) {
+				if((current_pos = VFS_LSEEK(fsp,fsp->fd,0,SEEK_CUR)) == -1) {
 					END_PROFILE(SMBlseek);
 					return(UNIXERROR(ERRDOS,ERRnoaccess));
 				}
@@ -2285,7 +2285,7 @@ int reply_lseek(connection_struct *conn, char *inbuf,char *outbuf, int size, int
 
 				SMB_STRUCT_STAT sbuf;
 
-				if(vfs_fstat(fsp,fsp->fd, &sbuf) == -1) {
+				if(VFS_FSTAT(fsp,fsp->fd, &sbuf) == -1) {
 					END_PROFILE(SMBlseek);
 					return(UNIXERROR(ERRDOS,ERRnoaccess));
 				}
@@ -2294,7 +2294,7 @@ int reply_lseek(connection_struct *conn, char *inbuf,char *outbuf, int size, int
 			}
  
 			if(current_pos < 0)
-				res = conn->vfs_ops.lseek(fsp,fsp->fd,0,SEEK_SET);
+				res = VFS_LSEEK(fsp,fsp->fd,0,SEEK_SET);
 		}
 
 		if(res == -1) {
@@ -2830,7 +2830,7 @@ NTSTATUS mkdir_internal(connection_struct *conn, pstring directory)
 	unix_convert(directory,conn,0,&bad_path,&sbuf);
 	
 	if (check_name(directory, conn))
-		ret = vfs_mkdir(conn,directory,unix_mode(conn,aDIR,directory));
+		ret = vfs_MkDir(conn,directory,unix_mode(conn,aDIR,directory));
 	
 	if (ret == -1) {
 		NTSTATUS nterr = set_bad_path_error(errno, bad_path);
@@ -2901,7 +2901,7 @@ static BOOL recursive_rmdir(connection_struct *conn, char *directory)
 		pstrcat(fullname, "/");
 		pstrcat(fullname, dname);
 
-		if(conn->vfs_ops.lstat(conn,fullname, &st) != 0) {
+		if(VFS_LSTAT(conn,fullname, &st) != 0) {
 			ret = True;
 			break;
 		}
@@ -2911,11 +2911,11 @@ static BOOL recursive_rmdir(connection_struct *conn, char *directory)
 				ret = True;
 				break;
 			}
-			if(vfs_rmdir(conn,fullname) != 0) {
+			if(VFS_RMDIR(conn,fullname) != 0) {
 				ret = True;
 				break;
 			}
-		} else if(vfs_unlink(conn,fullname) != 0) {
+		} else if(VFS_UNLINK(conn,fullname) != 0) {
 			ret = True;
 			break;
 		}
@@ -2932,7 +2932,7 @@ BOOL rmdir_internals(connection_struct *conn, char *directory)
 {
 	BOOL ok;
 
-	ok = (vfs_rmdir(conn,directory) == 0);
+	ok = (VFS_RMDIR(conn,directory) == 0);
 	if(!ok && ((errno == ENOTEMPTY)||(errno == EEXIST)) && lp_veto_files(SNUM(conn))) {
 		/* 
 		 * Check to see if the only thing in this directory are
@@ -2974,21 +2974,21 @@ BOOL rmdir_internals(connection_struct *conn, char *directory)
 					pstrcat(fullname, "/");
 					pstrcat(fullname, dname);
                      
-					if(conn->vfs_ops.lstat(conn,fullname, &st) != 0)
+					if(VFS_LSTAT(conn,fullname, &st) != 0)
 						break;
 					if(st.st_mode & S_IFDIR) {
 						if(lp_recursive_veto_delete(SNUM(conn))) {
 							if(recursive_rmdir(conn, fullname) != 0)
 								break;
 						}
-						if(vfs_rmdir(conn,fullname) != 0)
+						if(VFS_RMDIR(conn,fullname) != 0)
 							break;
-					} else if(vfs_unlink(conn,fullname) != 0)
+					} else if(VFS_UNLINK(conn,fullname) != 0)
 						break;
 				}
 				CloseDir(dirptr);
 				/* Retry the rmdir */
-				ok = (vfs_rmdir(conn,directory) == 0);
+				ok = (VFS_RMDIR(conn,directory) == 0);
 			} else {
 				CloseDir(dirptr);
 			}
@@ -3284,7 +3284,7 @@ directory = %s, newname = %s, newname_last_component = %s, is_8_3 = %d\n",
 			return NT_STATUS_OBJECT_NAME_COLLISION;
 		}
 
-		if(conn->vfs_ops.rename(conn,directory, newname) == 0) {
+		if(VFS_RENAME(conn,directory, newname) == 0) {
 			DEBUG(3,("rename_internals: succeeded doing rename on %s -> %s\n",
 				directory,newname));
 			return NT_STATUS_OK;	
@@ -3351,7 +3351,7 @@ directory = %s, newname = %s, newname_last_component = %s, is_8_3 = %d\n",
 					continue;
 				}
 				
-				if (!conn->vfs_ops.rename(conn,fname,destname))
+				if (!VFS_RENAME(conn,fname,destname))
 					count++;
 				DEBUG(3,("rename_internals: doing rename on %s -> %s\n",fname,destname));
 			}
@@ -3445,7 +3445,7 @@ static BOOL copy_file(char *src,char *dest1,connection_struct *conn, int ofun,
 	if (!target_is_directory && count)
 		ofun = FILE_EXISTS_OPEN;
 
-	if (vfs_stat(conn,dest,&sbuf2) == -1)
+	if (VFS_STAT(conn,dest,&sbuf2) == -1)
 		ZERO_STRUCTP(&sbuf2);
 
 	fsp2 = open_file_shared(conn,dest,&sbuf2,SET_DENY_MODE(DENY_NONE)|SET_OPEN_MODE(DOS_OPEN_WRONLY),
@@ -3457,7 +3457,7 @@ static BOOL copy_file(char *src,char *dest1,connection_struct *conn, int ofun,
 	}
 
 	if ((ofun&3) == 1) {
-		if(conn->vfs_ops.lseek(fsp2,fsp2->fd,0,SEEK_END) == -1) {
+		if(VFS_LSEEK(fsp2,fsp2->fd,0,SEEK_END) == -1) {
 			DEBUG(0,("copy_file: error - vfs lseek returned error %s\n", strerror(errno) ));
 			/*
 			 * Stop the copy from occurring.