diff options
author | Miles Bader <miles@gnu.org> | 1997-08-06 22:08:59 +0000 |
---|---|---|
committer | Miles Bader <miles@gnu.org> | 1997-08-06 22:08:59 +0000 |
commit | 80c437f2f35813085b86cc05987609ec36d18865 (patch) | |
tree | fdda379edf638d85d6bbf1bb9fb2934dba4cc1a6 /ftpfs/netfs.c | |
parent | f20a48afff1d935eea20aa2c02b5fd805e85663c (diff) |
Initial checkin
Diffstat (limited to 'ftpfs/netfs.c')
-rw-r--r-- | ftpfs/netfs.c | 441 |
1 files changed, 441 insertions, 0 deletions
diff --git a/ftpfs/netfs.c b/ftpfs/netfs.c new file mode 100644 index 00000000..2fdf8eec --- /dev/null +++ b/ftpfs/netfs.c @@ -0,0 +1,441 @@ +/* ftpfs interface to libnetfs + + Copyright (C) 1997 Free Software Foundation, Inc. + Written by Miles Bader <miles@gnu.ai.mit.edu> + This file is part of the GNU Hurd. + + The GNU Hurd 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, or (at + your option) any later version. + + The GNU Hurd 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., 59 Temple Place - Suite 330, Boston, MA 02111, USA. */ + +#include <stddef.h> +#include <stdlib.h> +#include <dirent.h> +#include <string.h> +#include <fcntl.h> + +#include <hurd/netfs.h> + +#include "ftpfs.h" +#include "ccache.h" + +/* Attempt to create a file named NAME in DIR for USER with MODE. Set *NODE + to the new node upon return. On any error, clear *NODE. *NODE should be + locked on success; no matter what, unlock DIR before returning. */ +error_t +netfs_attempt_create_file (struct iouser *user, struct node *dir, + char *name, mode_t mode, struct node **node) +{ + *node = 0; + mutex_unlock (&dir->lock); + return EOPNOTSUPP; +} + +/* Node NODE is being opened by USER, with FLAGS. NEWNODE is nonzero if we + just created this node. Return an error if we should not permit the open + to complete because of a permission restriction. */ +error_t +netfs_check_open_permissions (struct iouser *user, struct node *node, + int flags, int newnode) +{ + error_t err = ftpfs_refresh_node (node); + if (!err && (flags & O_READ)) + err = fshelp_access (&node->nn_stat, S_IREAD, user); + if (!err && (flags & O_WRITE)) + err = fshelp_access (&node->nn_stat, S_IWRITE, user); + if (!err && (flags & O_EXEC)) + err = fshelp_access (&node->nn_stat, S_IEXEC, user); + return err; +} + +/* This should attempt a utimes call for the user specified by CRED on node + NODE, to change the atime to ATIME and the mtime to MTIME. */ +error_t +netfs_attempt_utimes (struct iouser *cred, struct node *node, + struct timespec *atime, struct timespec *mtime) +{ + error_t err = ftpfs_refresh_node (node); + + if (! err) + err = fshelp_isowner (&node->nn_stat, cred); + + if (! err) + { + node->nn_stat.st_mtime = mtime->tv_sec; + node->nn_stat.st_mtime_usec = mtime->tv_nsec / 1000; + node->nn_stat.st_atime = atime->tv_sec; + node->nn_stat.st_atime_usec = atime->tv_nsec / 1000; + fshelp_touch (&node->nn_stat, TOUCH_CTIME, ftpfs_maptime); + } + + return err; +} + +/* Return the valid access types (bitwise OR of O_READ, O_WRITE, and O_EXEC) + in *TYPES for file NODE and user CRED. */ +error_t +netfs_report_access (struct iouser *cred, struct node *node, int *types) +{ + error_t err = ftpfs_refresh_node (node); + + if (! err) + { + *types = 0; + if (fshelp_access (&node->nn_stat, S_IREAD, cred) == 0) + *types |= O_READ; + if (fshelp_access (&node->nn_stat, S_IWRITE, cred) == 0) + *types |= O_WRITE; + if (fshelp_access (&node->nn_stat, S_IEXEC, cred) == 0) + *types |= O_EXEC; + } + + return err; +} + +/* Trivial definitions. */ + +/* Make sure that NP->nn_stat is filled with current information. CRED + identifies the user responsible for the operation. */ +error_t +netfs_validate_stat (struct node *node, struct iouser *cred) +{ + return ftpfs_refresh_node (node); +} + +/* This should sync the file NODE completely to disk, for the user CRED. If + WAIT is set, return only after sync is completely finished. */ +error_t +netfs_attempt_sync (struct iouser *cred, struct node *node, int wait) +{ + return 0; +} + +/* The granularity with which we allocate space to return our result. */ +#define DIRENTS_CHUNK_SIZE (8*1024) + +/* Returned directory entries are aligned to blocks this many bytes long. + Must be a power of two. */ +#define DIRENT_ALIGN 4 +#define DIRENT_NAME_OFFS offsetof (struct dirent, d_name) + +/* Length is structure before the name + the name + '\0', all + padded to a four-byte alignment. */ +#define DIRENT_LEN(name_len) \ + ((DIRENT_NAME_OFFS + (name_len) + 1 + (DIRENT_ALIGN - 1)) \ + & ~(DIRENT_ALIGN - 1)) + +/* Fetch a directory, as for netfs_get_dirents. */ +static error_t +get_dirents (struct ftpfs_dir *dir, + int first_entry, int max_entries, char **data, + mach_msg_type_number_t *data_len, + vm_size_t max_data_len, int *data_entries) +{ + struct ftpfs_dir_entry *e; + error_t err = 0; + + if (! dir) + return ENOTDIR; + + e = dir->ordered; + + /* Find the first entry. */ + while (first_entry-- > 0) + if (! e) + { + max_entries = 0; + break; + } + else + e = e->ordered_next; + + if (max_entries != 0) + { + size_t size = + (max_data_len == 0 || max_data_len > DIRENTS_CHUNK_SIZE + ? DIRENTS_CHUNK_SIZE + : max_data_len); + + err = vm_allocate (mach_task_self (), (vm_address_t *) data, size, 1); + + if (! err) + { + char *p = *data; + int count = 0; + + /* See how much space we need for the result. */ + while ((max_entries == -1 || count < max_entries) && e) + { + struct dirent hdr; + size_t name_len = strlen (e->name); + size_t sz = DIRENT_LEN (name_len); + int entry_type = + e->stat_timestamp ? IFTODT (e->stat.st_mode) : DT_UNKNOWN; + + if ((p - *data) + sz > size) + if (max_data_len > 0) + break; + else + /* Try to grow our return buffer. */ + { + vm_address_t extension = (vm_address_t)(*data + size); + err = vm_allocate (mach_task_self (), &extension, + DIRENTS_CHUNK_SIZE, 0); + if (err) + break; + size += DIRENTS_CHUNK_SIZE; + } + + hdr.d_namlen = name_len; + hdr.d_fileno = e->stat.st_ino; + hdr.d_reclen = sz; + hdr.d_type = entry_type; + + memcpy (p, &hdr, DIRENT_NAME_OFFS); + strcpy (p + DIRENT_NAME_OFFS, e->name); + p += sz; + + count++; + e = e->ordered_next; + } + + if (err) + vm_deallocate (mach_task_self (), (vm_address_t)*data, size); + else + { + vm_address_t alloc_end = (vm_address_t)(*data + size); + vm_address_t real_end = round_page (p); + if (alloc_end > real_end) + vm_deallocate (mach_task_self (), + real_end, alloc_end - real_end); + *data_len = p - *data; + *data_entries = count; + } + } + } + else + { + *data_len = 0; + *data_entries = 0; + } + + return err; +} + +error_t +netfs_get_dirents (struct iouser *cred, struct node *dir, + int first_entry, int max_entries, char **data, + mach_msg_type_number_t *data_len, + vm_size_t max_data_len, int *data_entries) +{ + error_t err = ftpfs_refresh_node (dir); + + if (! err) + if (dir->nn->dir) + { + err = ftpfs_dir_refresh (dir->nn->dir); + if (! err) + err = get_dirents (dir->nn->dir, first_entry, max_entries, + data, data_len, max_entries, data_entries); + } + else + err = ENOTDIR; + + return err; +} + +/* Lookup NAME in DIR for USER; set *NODE to the found name upon return. If + the name was not found, then return ENOENT. On any error, clear *NODE. + (*NODE, if found, should be locked, this call should unlock DIR no matter + what.) */ +error_t netfs_attempt_lookup (struct iouser *user, struct node *dir, + char *name, struct node **node) +{ + error_t err = ftpfs_refresh_node (dir); + if (! err) + err = ftpfs_dir_lookup (dir->nn->dir, name, node); + return err; +} + +/* Delete NAME in DIR for USER. */ +error_t netfs_attempt_unlink (struct iouser *user, struct node *dir, + char *name) +{ + return EROFS; +} + +/* Note that in this one call, neither of the specific nodes are locked. */ +error_t netfs_attempt_rename (struct iouser *user, struct node *fromdir, + char *fromname, struct node *todir, + char *toname, int excl) +{ + return EROFS; +} + +/* Attempt to create a new directory named NAME in DIR for USER with mode + MODE. */ +error_t netfs_attempt_mkdir (struct iouser *user, struct node *dir, + char *name, mode_t mode) +{ + return EROFS; +} + +/* Attempt to remove directory named NAME in DIR for USER. */ +error_t netfs_attempt_rmdir (struct iouser *user, + struct node *dir, char *name) +{ + return EROFS; +} + +/* This should attempt a chmod call for the user specified by CRED on node + NODE, to change the owner to UID and the group to GID. */ +error_t netfs_attempt_chown (struct iouser *cred, struct node *node, + uid_t uid, uid_t gid) +{ + return EROFS; +} + +/* This should attempt a chauthor call for the user specified by CRED on node + NODE, to change the author to AUTHOR. */ +error_t netfs_attempt_chauthor (struct iouser *cred, struct node *node, + uid_t author) +{ + return EROFS; +} + +/* This should attempt a chmod call for the user specified by CRED on node + NODE, to change the mode to MODE. Unlike the normal Unix and Hurd meaning + of chmod, this function is also used to attempt to change files into other + types. If such a transition is attempted which is impossible, then return + EOPNOTSUPP. */ +error_t netfs_attempt_chmod (struct iouser *cred, struct node *node, + mode_t mode) +{ + return EROFS; +} + +/* Attempt to turn NODE (user CRED) into a symlink with target NAME. */ +error_t netfs_attempt_mksymlink (struct iouser *cred, struct node *node, + char *name) +{ + return EROFS; +} + +/* Attempt to turn NODE (user CRED) into a device. TYPE is either S_IFBLK or + S_IFCHR. */ +error_t netfs_attempt_mkdev (struct iouser *cred, struct node *node, + mode_t type, dev_t indexes) +{ + return EROFS; +} + +/* Attempt to set the passive translator record for FILE to ARGZ (of length + ARGZLEN) for user CRED. */ +error_t netfs_set_translator (struct iouser *cred, struct node *node, + char *argz, size_t argzlen) +{ + return EROFS; +} + +/* This should attempt a chflags call for the user specified by CRED on node + NODE, to change the flags to FLAGS. */ +error_t netfs_attempt_chflags (struct iouser *cred, struct node *node, + int flags) +{ + return EROFS; +} + +/* This should attempt to set the size of the file NODE (for user CRED) to + SIZE bytes long. */ +error_t netfs_attempt_set_size (struct iouser *cred, struct node *node, + off_t size) +{ + return EROFS; +} + +/* This should attempt to fetch filesystem status information for the remote + filesystem, for the user CRED. */ +error_t netfs_attempt_statfs (struct iouser *cred, struct node *node, + struct statfs *st) +{ + return EOPNOTSUPP; +} + +/* This should sync the entire remote filesystem. If WAIT is set, return + only after sync is completely finished. */ +error_t netfs_attempt_syncfs (struct iouser *cred, int wait) +{ + return 0; +} + +/* Create a link in DIR with name NAME to FILE for USER. Note that neither + DIR nor FILE are locked. If EXCL is set, do not delete the target, but + return EEXIST if NAME is already found in DIR. */ +error_t netfs_attempt_link (struct iouser *user, struct node *dir, + struct node *file, char *name, int excl) +{ + return EROFS; +} + +/* Attempt to create an anonymous file related to DIR for USER with MODE. + Set *NODE to the returned file upon success. No matter what, unlock DIR. */ +error_t netfs_attempt_mkfile (struct iouser *user, struct node *dir, + mode_t mode, struct node **node) +{ + return EROFS; +} + +/* Read the contents of NODE (a symlink), for USER, into BUF. */ +error_t netfs_attempt_readlink (struct iouser *user, struct node *node, char *buf) +{ + error_t err = ftpfs_refresh_node (node); + if (! err) + { + struct ftpfs_dir_entry *e = node->nn->dir_entry; + if (e) + bcopy (e->symlink_target, buf, node->nn_stat.st_size); + else + err = EINVAL; + } + return err; +} + +/* Read from the file NODE for user CRED starting at OFFSET and continuing for + up to *LEN bytes. Put the data at DATA. Set *LEN to the amount + successfully read upon return. */ +error_t netfs_attempt_read (struct iouser *cred, struct node *node, + off_t offset, size_t *len, void *data) +{ + error_t err = 0; + + if (! node->nn->contents) + err = ccache_create (node, &node->nn->contents); + if (! err) + { + if (*len > node->nn_stat.st_size - offset) + *len = node->nn_stat.st_size - offset; + if (*len > 0) + err = ccache_read (node->nn->contents, offset, *len, data); + } + + return err; +} + +/* Write to the file NODE for user CRED starting at OFSET and continuing for up + to *LEN bytes from DATA. Set *LEN to the amount seccessfully written upon + return. */ +error_t netfs_attempt_write (struct iouser *cred, struct node *node, + off_t offset, size_t *len, void *data) +{ + return EROFS; +} |