+}
+
+struct v_inode*
+vfs_i_find(struct v_superblock* sb, u32_t i_id)
+{
+ struct hbucket* slot;
+ struct v_inode *pos, *n, *found = NULL;
+
+ cache_atomic_read(&sb->i_cache,
+ {
+ slot = &sb->i_cache.pool[i_id & VFS_HASH_MASK];
+
+ hashtable_bucket_foreach(slot, pos, n, hash_list)
+ {
+ if (pos->id != i_id) {
+ continue;
+ }
+
+ lru_use_one(inode_lru, &pos->lru);
+ found = pos;
+ break;
+ }
+ });
+
+ return found;
+}
+
+void
+vfs_i_addhash(struct v_inode* inode)
+{
+ vfs_vncache_add(inode_cache(inode), inode->id, &inode->hash_list);
+}
+
+struct v_inode*
+vfs_i_alloc(struct v_superblock* sb)
+{
+ assert(sb->ops.init_inode);
+
+ struct v_inode* inode;
+ if (!(inode = cake_grab(inode_pile))) {
+ lru_evict_half(inode_lru);
+ if (!(inode = cake_grab(inode_pile))) {
+ return NULL;
+ }
+ }
+
+ memset(inode, 0, sizeof(*inode));
+ mutex_init(&inode->lock);
+ llist_init_head(&inode->xattrs);
+ llist_init_head(&inode->aka_dnodes);
+
+ sb->ops.init_inode(sb, inode);
+
+ inode->ctime = clock_unixtime();
+ inode->atime = inode->ctime;
+ inode->mtime = inode->ctime;
+
+ vfs_i_assign_sb(inode, sb);
+ lru_use_one(inode_lru, &inode->lru);
+
+ return inode;
+}
+
+void
+vfs_i_free(struct v_inode* inode)
+{
+ if (inode->pg_cache) {
+ pcache_release(inode->pg_cache);
+ vfree(inode->pg_cache);
+ }
+
+ // we don't need to sync inode.
+ // If an inode can be free, then it must be properly closed.
+ // Hence it must be synced already!
+ if (inode->destruct) {
+ inode->destruct(inode);
+ }
+
+ vfs_sb_unref(inode->sb);
+
+ hlist_delete(&inode->hash_list);
+ lru_remove(inode_lru, &inode->lru);
+
+ cake_release(inode_pile, inode);
+}
+
+/* ---- System call definition and support ---- */
+
+// make a new name when not exists
+#define FLOC_MAYBE_MKNAME 1
+
+// name must be non-exist and made.
+#define FLOC_MKNAME 2
+
+// no follow symlink
+#define FLOC_NOFOLLOW 4
+
+int
+vfs_getfd(int fd, struct v_fd** fd_s)
+{
+ struct v_fdtable* fdtab;
+
+ if (!TEST_FD(fd)) {
+ return EBADF;
+ }
+
+ fdtab = __current->fdtable;
+
+ lock_fdtable(fdtab);
+ *fd_s = __current->fdtable->fds[fd];
+ unlock_fdtable(fdtab);
+
+ return !*fd_s ? EBADF : 0;
+}
+
+static int
+__vfs_mknod(struct v_inode* parent, struct v_dnode* dnode,
+ unsigned int itype, dev_t* dev)
+{
+ int errno;
+
+ errno = parent->ops->create(parent, dnode, itype);
+ if (errno) {
+ return errno;
+ }
+
+ return 0;
+}
+
+struct file_locator {
+ struct v_dnode* dir;
+ struct v_dnode* file;
+ bool fresh;
+};
+
+/**
+ * @brief unlock the file locator (floc) if possible.
+ * If the file to be located if not exists, and
+ * any FLOC_*MKNAME flag is set, then the parent
+ * dnode will be locked until the file has been properly
+ * finalised by subsequent logic.
+ *
+ * @param floc
+ */
+static inline void
+__floc_try_unlock(struct file_locator* floc)
+{
+ if (floc->fresh) {
+ assert(floc->dir);
+ unlock_dnode(floc->dir);
+ }
+}
+
+static int
+__vfs_try_locate_file(const char* path,
+ struct file_locator* floc,
+ int options)
+{
+ char name_str[VFS_NAME_MAXLEN];
+ struct v_dnode *fdir, *file;
+ struct hstr name = HSTR(name_str, 0);
+ int errno, woption = 0;
+
+ if ((options & FLOC_NOFOLLOW)) {
+ woption |= VFS_WALK_NOFOLLOW;
+ options &= ~FLOC_NOFOLLOW;
+ }
+
+ floc->fresh = false;
+ name_str[0] = 0;
+ errno = vfs_walk_proc(path, &fdir, &name, woption | VFS_WALK_PARENT);
+ if (errno) {
+ return errno;
+ }
+
+ lock_dnode(fdir);
+
+ errno = vfs_walk(fdir, name.value, &file, NULL, woption);
+
+ if (errno && errno != ENOENT) {
+ goto error;
+ }
+
+ if (!errno && (options & FLOC_MKNAME)) {
+ errno = EEXIST;
+ goto error;
+ }
+
+ if (!errno) {
+ // the file present, no need to hold the directory lock
+ unlock_dnode(fdir);
+ goto done;
+ }
+
+ // errno == ENOENT
+ if (!options) {
+ goto error;
+ }
+
+ errno = vfs_check_writable(fdir);
+ if (errno) {
+ goto error;
+ }
+
+ floc->fresh = true;
+
+ file = vfs_d_alloc(fdir, &name);
+
+ if (!file) {
+ errno = ENOMEM;
+ goto error;
+ }
+
+ vfs_dcache_add(fdir, file);
+
+done:
+ floc->dir = fdir;
+ floc->file = file;
+
+ return errno;
+
+error:
+ unlock_dnode(fdir);
+ return errno;
+}
+
+
+static bool
+__check_unlinkable(struct v_dnode* dnode)
+{
+ int acl;
+ bool wr_self, wr_parent;
+ struct v_dnode* parent;
+
+ parent = dnode->parent;
+ acl = dnode->inode->acl;
+
+ wr_self = check_allow_write(dnode->inode);
+ wr_parent = check_allow_write(parent->inode);
+
+ if (!fsacl_test(acl, svtx)) {
+ return wr_self;
+ }
+
+ if (current_euid() == dnode->inode->uid) {
+ return true;
+ }
+
+ return wr_self && wr_parent;
+}
+
+int
+vfs_do_open(const char* path, int options)
+{
+ int errno, fd, loptions = 0;
+ struct v_dnode *dentry, *file;
+ struct v_file* ofile = NULL;
+ struct file_locator floc;
+ struct v_inode* inode;
+
+ if ((options & FO_CREATE)) {
+ loptions |= FLOC_MAYBE_MKNAME;
+ } else if ((options & FO_NOFOLLOW)) {
+ loptions |= FLOC_NOFOLLOW;
+ }
+
+ errno = __vfs_try_locate_file(path, &floc, loptions);
+
+ if (errno || (errno = vfs_alloc_fdslot(&fd))) {
+ return errno;
+ }
+
+ file = floc.file;
+ dentry = floc.dir;
+
+ if (floc.fresh) {
+ errno = __vfs_mknod(dentry->inode, file, VFS_IFFILE, NULL);
+ if (errno) {
+ vfs_d_free(file);
+ __floc_try_unlock(&floc);
+ return errno;
+ }
+
+ __floc_try_unlock(&floc);
+ }
+
+
+ if ((errno = vfs_open(file, &ofile))) {
+ return errno;
+ }
+
+ inode = ofile->inode;
+ lock_inode(inode);
+
+ struct v_fd* fd_s = cake_grab(fd_pile);
+ memset(fd_s, 0, sizeof(*fd_s));
+
+ if ((options & O_TRUNC)) {
+ file->inode->fsize = 0;
+ }
+
+ if (vfs_get_dtype(inode->itype) == DT_DIR) {
+ ofile->f_pos = 0;
+ }
+
+ fd_s->file = ofile;
+ fd_s->flags = options;
+ __current->fdtable->fds[fd] = fd_s;
+
+ unlock_inode(inode);
+
+ return fd;
+}
+
+__DEFINE_LXSYSCALL2(int, open, const char*, path, int, options)
+{
+ int errno = vfs_do_open(path, options);
+ return DO_STATUS_OR_RETURN(errno);
+}
+
+__DEFINE_LXSYSCALL1(int, close, int, fd)
+{
+ struct v_fd* fd_s;
+ int errno = 0;
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done_err;
+ }
+
+ if ((errno = vfs_close(fd_s->file))) {
+ goto done_err;
+ }
+
+ cake_release(fd_pile, fd_s);
+ __current->fdtable->fds[fd] = 0;
+
+done_err:
+ return DO_STATUS(errno);
+}
+
+void
+__vfs_readdir_callback(struct dir_context* dctx,
+ const char* name,
+ const int len,
+ const int dtype)
+{
+ struct lx_dirent* dent = (struct lx_dirent*)dctx->cb_data;
+ strncpy(dent->d_name, name, MIN(len, DIRENT_NAME_MAX_LEN));
+ dent->d_nlen = len;
+ dent->d_type = dtype;
+}
+
+__DEFINE_LXSYSCALL2(int, sys_readdir, int, fd, struct lx_dirent*, dent)
+{
+ struct v_fd* fd_s;
+ int errno;
+
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_inode* inode = fd_s->file->inode;
+
+ lock_inode(inode);
+
+ if (!check_directory_node(inode)) {
+ errno = ENOTDIR;
+ goto unlock;
+ }
+
+ if (!check_allow_read(inode)) {
+ errno = EPERM;
+ goto unlock;
+ }
+
+ struct dir_context dctx = (struct dir_context) {
+ .cb_data = dent,
+ .read_complete_callback = __vfs_readdir_callback
+ };
+
+ if ((errno = fd_s->file->ops->readdir(fd_s->file, &dctx)) != 1) {
+ goto unlock;
+ }
+
+ dent->d_offset++;
+ fd_s->file->f_pos++;
+
+unlock:
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS_OR_RETURN(errno);
+}
+
+__DEFINE_LXSYSCALL3(int, read, int, fd, void*, buf, size_t, count)
+{
+ int errno = 0;
+ struct v_fd* fd_s;
+ struct v_inode* inode;
+
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_file* file = fd_s->file;
+ if (check_directory_node(file->inode)) {
+ errno = EISDIR;
+ goto done;
+ }
+
+ if (!check_allow_read(file->inode)) {
+ errno = EPERM;
+ goto done;
+ }
+
+ inode = file->inode;
+ lock_inode(inode);
+
+ __vfs_touch_inode(inode, INODE_ACCESSED);
+
+ if (check_seqdev_node(inode) || (fd_s->flags & FO_DIRECT)) {
+ errno = file->ops->read(inode, buf, count, file->f_pos);
+ } else {
+ errno = pcache_read(inode, buf, count, file->f_pos);
+ }
+
+ if (errno > 0) {
+ file->f_pos += errno;
+ unlock_inode(inode);
+ return errno;
+ }
+
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL3(int, write, int, fd, void*, buf, size_t, count)
+{
+ int errno = 0;
+ struct v_fd* fd_s;
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_inode* inode;
+ struct v_file* file = fd_s->file;
+
+ if ((errno = vfs_check_writable(file->dnode))) {
+ goto done;
+ }
+
+ if (check_directory_node(file->inode)) {
+ errno = EISDIR;
+ goto done;
+ }
+
+ inode = file->inode;
+ lock_inode(inode);
+
+ __vfs_touch_inode(inode, INODE_MODIFY);
+ if ((fd_s->flags & O_APPEND)) {
+ file->f_pos = inode->fsize;
+ }
+
+ if (check_seqdev_node(inode) || (fd_s->flags & FO_DIRECT)) {
+ errno = file->ops->write(inode, buf, count, file->f_pos);
+ } else {
+ errno = pcache_write(inode, buf, count, file->f_pos);
+ }
+
+ if (errno > 0) {
+ file->f_pos += errno;
+ inode->fsize = MAX(inode->fsize, file->f_pos);
+
+ unlock_inode(inode);
+ return errno;
+ }
+
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL3(int, lseek, int, fd, int, offset, int, options)
+{
+ int errno = 0;
+ struct v_fd* fd_s;
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_file* file = fd_s->file;
+ struct v_inode* inode = file->inode;
+
+ if (!file->ops->seek) {
+ errno = ENOTSUP;
+ goto done;
+ }
+
+ if (!check_allow_read(inode)) {
+ errno = EPERM;
+ goto done;
+ }
+
+ lock_inode(inode);
+
+ int overflow = 0;
+ int fpos = file->f_pos;
+
+ if (vfs_get_dtype(inode->itype) == DT_DIR) {
+ options = (options != FSEEK_END) ? options : FSEEK_SET;
+ }
+
+ switch (options) {
+ case FSEEK_CUR:
+ overflow = sadd_of((int)file->f_pos, offset, &fpos);
+ break;
+ case FSEEK_END:
+ overflow = sadd_of((int)inode->fsize, offset, &fpos);
+ break;
+ case FSEEK_SET:
+ fpos = offset;
+ break;
+ }
+
+ if (overflow) {
+ errno = EOVERFLOW;
+ }
+ else {
+ errno = file->ops->seek(file, fpos);
+ }
+
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+int
+vfs_get_path(struct v_dnode* dnode, char* buf, size_t size, int depth)
+{
+ if (!dnode) {
+ return 0;
+ }
+
+ if (depth > 64) {
+ return ENAMETOOLONG;
+ }
+
+ size_t len = 0;
+
+ if (dnode->parent != dnode) {
+ len = vfs_get_path(dnode->parent, buf, size, depth + 1);
+ }
+
+ if (len >= size) {
+ return len;
+ }
+
+ if (!len || buf[len - 1] != VFS_PATH_DELIM) {
+ buf[len++] = VFS_PATH_DELIM;
+ }
+
+ size_t cpy_size = MIN(dnode->name.len, size - len);
+ strncpy(buf + len, dnode->name.value, cpy_size);
+ len += cpy_size;
+
+ return len;
+}
+
+int
+vfs_readlink(struct v_dnode* dnode, char* buf, size_t size)
+{
+ const char* link;
+ struct v_inode* inode = dnode->inode;
+
+ if (!check_symlink_node(inode)) {
+ return EINVAL;
+ }
+
+ if (!inode->ops->read_symlink) {
+ return ENOTSUP;
+ }
+
+ if (!check_allow_read(inode)) {
+ return EPERM;
+ }
+
+ lock_inode(inode);
+
+ int errno = inode->ops->read_symlink(inode, &link);
+ if (errno >= 0) {
+ strncpy(buf, link, MIN(size, (size_t)errno));
+ }
+
+ unlock_inode(inode);
+ return errno;
+}
+
+int
+vfs_get_dtype(int itype)
+{
+ int dtype = DT_FILE;
+ if (check_itype(itype, VFS_IFSYMLINK)) {
+ dtype |= DT_SYMLINK;
+ }
+
+ if (check_itype(itype, VFS_IFDIR)) {
+ dtype |= DT_DIR;
+ return dtype;
+ }
+
+ // TODO other types
+
+ return dtype;
+}
+
+struct v_fdtable*
+fdtable_create()
+{
+ struct v_fdtable* fdtab;
+
+ fdtab = vzalloc(sizeof(struct v_fdtable));
+ mutex_init(&fdtab->lock);
+
+ return fdtab;
+}
+
+void
+fdtable_copy(struct v_fdtable* dest, struct v_fdtable* src)
+{
+ lock_fdtable(dest);
+ lock_fdtable(src);
+
+ for (size_t i = 0; i < VFS_MAX_FD; i++) {
+ struct v_fd* fd = src->fds[i];
+ if (!fd)
+ continue;
+ vfs_dup_fd(fd, &dest->fds[i]);
+ }
+
+ unlock_fdtable(dest);
+ unlock_fdtable(src);
+}
+
+void
+fdtable_free(struct v_fdtable* table)
+{
+ assert(!mutex_on_hold(&table->lock));
+
+ vfree(table);
+}
+
+__DEFINE_LXSYSCALL3(int, realpathat, int, fd, char*, buf, size_t, size)
+{
+ int errno;
+ struct v_fd* fd_s;
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_dnode* dnode;
+
+ dnode = fd_s->file->dnode;
+
+ lock_dnode(dnode);
+ errno = vfs_get_path(dnode, buf, size, 0);
+ unlock_dnode(dnode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL3(int, readlink, const char*, path, char*, buf, size_t, size)
+{
+ int errno;
+ struct v_dnode* dnode;
+ if (!(errno = vfs_walk_proc(path, &dnode, NULL, VFS_WALK_NOFOLLOW))) {
+ errno = vfs_readlink(dnode, buf, size);
+ }
+
+ if (errno >= 0) {
+ return errno;
+ }
+
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL4(
+ int, readlinkat, int, dirfd, const char*, pathname, char*, buf, size_t, size)
+{
+ int errno;
+ struct v_fd* fd_s;
+ if ((errno = vfs_getfd(dirfd, &fd_s))) {
+ goto done;
+ }
+
+ pathname = pathname ? pathname : "";
+
+ struct v_dnode* dnode;
+ if (!(errno = vfs_walk(
+ fd_s->file->dnode, pathname, &dnode, NULL, VFS_WALK_NOFOLLOW))) {
+ errno = vfs_readlink(fd_s->file->dnode, buf, size);
+ }
+
+ if (errno >= 0) {
+ return errno;
+ }
+
+done:
+ return DO_STATUS(errno);
+}
+
+/*
+ NOTE
+ When we perform operation that could affect the layout of
+ directory (i.e., rename, mkdir, rmdir). We must lock the parent dir
+ whenever possible. This will blocking any ongoing path walking to reach
+ it hence avoid any partial state.
+*/
+
+__DEFINE_LXSYSCALL1(int, rmdir, const char*, pathname)
+{
+ int errno;
+ struct v_dnode* dnode;
+ if ((errno = vfs_walk_proc(pathname, &dnode, NULL, 0))) {
+ return DO_STATUS(errno);
+ }
+
+ lock_dnode(dnode);
+
+ if (!__check_unlinkable(dnode)) {
+ errno = EPERM;
+ goto done;
+ }
+
+ if ((errno = vfs_check_writable(dnode))) {
+ goto done;
+ }
+
+ if ((dnode->super_block->fs->types & FSTYPE_ROFS)) {
+ errno = EROFS;
+ goto done;
+ }
+
+ if (dnode->ref_count > 1 || dnode->inode->open_count) {
+ errno = EBUSY;
+ goto done;
+ }
+
+ if (!llist_empty(&dnode->children)) {
+ errno = ENOTEMPTY;
+ goto done;
+ }
+
+ struct v_dnode* parent = dnode->parent;
+
+ if (!parent) {
+ errno = EINVAL;
+ goto done;
+ }
+
+ lock_dnode(parent);
+ lock_inode(parent->inode);
+
+ if (check_directory_node(dnode->inode)) {
+ errno = parent->inode->ops->rmdir(parent->inode, dnode);
+ if (!errno) {
+ vfs_dcache_remove(dnode);
+ }
+ } else {
+ errno = ENOTDIR;
+ }
+
+ unlock_inode(parent->inode);
+ unlock_dnode(parent);
+
+done:
+ unlock_dnode(dnode);
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL1(int, mkdir, const char*, path)
+{
+ int errno;
+ struct hstr name;
+ struct v_inode* inode;
+ struct v_dnode *parent, *dir;
+ char name_value[VFS_NAME_MAXLEN];
+
+ name = HHSTR(name_value, 0, 0);
+
+ if ((errno = vfs_walk_proc(path, &parent, &name, VFS_WALK_PARENT))) {
+ goto done;
+ }
+
+ if (!(errno = vfs_walk(parent, name_value, &dir, NULL, 0))) {
+ errno = EEXIST;
+ goto done;
+ }
+
+ if ((errno = vfs_check_writable(parent))) {
+ goto done;
+ }
+
+ if (!(dir = vfs_d_alloc(parent, &name))) {
+ errno = ENOMEM;
+ goto done;
+ }
+
+ inode = parent->inode;
+
+ lock_dnode(parent);
+ lock_inode(inode);
+
+ if ((parent->super_block->fs->types & FSTYPE_ROFS)) {
+ errno = ENOTSUP;
+ } else if (!inode->ops->mkdir) {
+ errno = ENOTSUP;
+ } else if (!check_directory_node(inode)) {
+ errno = ENOTDIR;
+ } else if (!(errno = inode->ops->mkdir(inode, dir))) {
+ vfs_dcache_add(parent, dir);
+ goto cleanup;
+ }
+
+ vfs_d_free(dir);
+
+cleanup:
+ unlock_inode(inode);
+ unlock_dnode(parent);
+done:
+ return DO_STATUS(errno);
+}
+
+static int
+__vfs_do_unlink(struct v_dnode* dnode)
+{
+ int errno;
+ struct v_inode* inode = dnode->inode;
+
+ if (dnode->ref_count > 1) {
+ return EBUSY;
+ }
+
+ if (!__check_unlinkable(dnode)) {
+ return EPERM;
+ }
+
+ if ((errno = vfs_check_writable(dnode))) {
+ return errno;
+ }
+
+ lock_inode(inode);
+
+ if (inode->open_count) {
+ errno = EBUSY;
+ } else if (!check_directory_node(inode)) {
+ errno = inode->ops->unlink(inode, dnode);
+ if (!errno) {
+ vfs_d_free(dnode);
+ }
+ } else {
+ errno = EISDIR;
+ }
+
+ unlock_inode(inode);
+
+ return errno;
+}
+
+__DEFINE_LXSYSCALL1(int, unlink, const char*, pathname)
+{
+ int errno;
+ struct v_dnode* dnode;
+ if ((errno = vfs_walk_proc(pathname, &dnode, NULL, 0))) {
+ goto done;
+ }
+
+ errno = __vfs_do_unlink(dnode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(int, unlinkat, int, fd, const char*, pathname)
+{
+ int errno;
+ struct v_fd* fd_s;
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ struct v_dnode* dnode;
+ if (!(errno = vfs_walk(fd_s->file->dnode, pathname, &dnode, NULL, 0))) {
+ errno = __vfs_do_unlink(dnode);
+ }
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(int, link, const char*, oldpath, const char*, newpath)
+{
+ int errno;
+ struct file_locator floc;
+ struct v_dnode *to_link, *name_file;
+
+ errno = __vfs_try_locate_file(oldpath, &floc, 0);
+ if (errno) {
+ goto done;
+ }
+
+ __floc_try_unlock(&floc);
+
+ to_link = floc.file;
+ errno = __vfs_try_locate_file(newpath, &floc, FLOC_MKNAME);
+ if (!errno) {
+ goto done;
+ }
+
+ name_file = floc.file;
+ errno = vfs_link(to_link, name_file);
+ if (errno) {
+ vfs_d_free(name_file);
+ }
+
+done:
+ __floc_try_unlock(&floc);
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL1(int, fsync, int, fildes)
+{
+ int errno;
+ struct v_fd* fd_s;
+
+ if (!(errno = vfs_getfd(fildes, &fd_s))) {
+ errno = vfs_fsync(fd_s->file);
+ }
+
+ return DO_STATUS(errno);
+}
+
+int
+vfs_dup_fd(struct v_fd* old, struct v_fd** new)
+{
+ int errno = 0;
+ struct v_fd* copied = cake_grab(fd_pile);
+
+ memcpy(copied, old, sizeof(struct v_fd));
+
+ vfs_ref_file(old->file);
+
+ *new = copied;
+
+ return errno;
+}
+
+int
+vfs_dup2(int oldfd, int newfd)
+{
+ int errno;
+ struct v_fdtable* fdtab;
+ struct v_fd *oldfd_s, *newfd_s;
+
+ if (newfd == oldfd) {
+ return newfd;
+ }
+
+ if ((errno = vfs_getfd(oldfd, &oldfd_s))) {
+ goto done;
+ }
+
+ if (!TEST_FD(newfd)) {
+ errno = EBADF;
+ goto done;
+ }
+
+ fdtab = __current->fdtable;
+ lock_fdtable(fdtab);
+
+ newfd_s = fdtab->fds[newfd];
+ if (newfd_s && (errno = vfs_close(newfd_s->file))) {
+ goto unlock_and_done;
+ }
+
+ if ((errno = vfs_dup_fd(oldfd_s, &newfd_s))) {
+ goto unlock_and_done;
+ }
+
+ fdtab->fds[newfd] = newfd_s;
+
+ unlock_fdtable(fdtab);
+ return newfd;
+
+unlock_and_done:
+ unlock_fdtable(fdtab);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(int, dup2, int, oldfd, int, newfd)
+{
+ return vfs_dup2(oldfd, newfd);
+}
+
+__DEFINE_LXSYSCALL1(int, dup, int, oldfd)
+{
+ int errno, newfd;
+ struct v_fd *oldfd_s, *newfd_s;
+ if ((errno = vfs_getfd(oldfd, &oldfd_s))) {
+ goto done;
+ }
+
+ if (!(errno = vfs_alloc_fdslot(&newfd)) &&
+ !(errno = vfs_dup_fd(oldfd_s, &newfd_s))) {
+ __current->fdtable->fds[newfd] = newfd_s;
+ return newfd;
+ }
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(
+ int, symlink, const char*, pathname, const char*, link_target)
+{
+ int errno;
+ struct file_locator floc;
+ struct v_dnode *file;
+ struct v_inode *f_ino;
+
+ errno = __vfs_try_locate_file(pathname, &floc, FLOC_MKNAME);
+ if (errno) {
+ goto done;
+ }
+
+ file = floc.file;
+ errno = __vfs_mknod(floc.dir->inode, file, VFS_IFSYMLINK, NULL);
+ if (errno) {
+ vfs_d_free(file);
+ goto done;
+ }
+
+ f_ino = file->inode;
+
+ assert(f_ino);
+
+ errno = vfs_check_writable(file);
+ if (errno) {
+ goto done;
+ }
+
+ if (!f_ino->ops->set_symlink) {
+ errno = ENOTSUP;
+ goto done;
+ }
+
+ lock_inode(f_ino);
+
+ errno = f_ino->ops->set_symlink(f_ino, link_target);
+
+ unlock_inode(f_ino);
+
+done:
+ __floc_try_unlock(&floc);
+ return DO_STATUS(errno);
+}
+
+static int
+vfs_do_chdir_nolock(struct proc_info* proc, struct v_dnode* dnode)
+{
+ if (!check_directory_node(dnode->inode)) {
+ return ENOTDIR;
+ }
+
+ if (proc->cwd) {
+ vfs_unref_dnode(proc->cwd);
+ }
+
+ vfs_ref_dnode(dnode);
+ proc->cwd = dnode;
+
+ return 0;
+}
+
+static int
+vfs_do_chdir(struct proc_info* proc, struct v_dnode* dnode)
+{
+ int errno = 0;
+
+ lock_dnode(dnode);
+
+ errno = vfs_do_chdir_nolock(proc, dnode);
+
+ unlock_dnode(dnode);
+
+ return errno;
+}
+
+__DEFINE_LXSYSCALL1(int, chdir, const char*, path)
+{
+ struct v_dnode* dnode;
+ int errno = 0;
+
+ if ((errno = vfs_walk_proc(path, &dnode, NULL, 0))) {
+ goto done;
+ }
+
+ errno = vfs_do_chdir((struct proc_info*)__current, dnode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL1(int, fchdir, int, fd)
+{
+ struct v_fd* fd_s;
+ int errno = 0;
+
+ if ((errno = vfs_getfd(fd, &fd_s))) {
+ goto done;
+ }
+
+ errno = vfs_do_chdir((struct proc_info*)__current, fd_s->file->dnode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+
+__DEFINE_LXSYSCALL1(int, chroot, const char*, path)
+{
+ int errno;
+ struct v_dnode* dnode;
+ if ((errno = vfs_walk_proc(path, &dnode, NULL, 0))) {
+ return errno;
+ }
+
+ lock_dnode(dnode);
+
+ errno = vfs_do_chdir_nolock(__current, dnode);
+ if (errno) {
+ unlock_dnode(dnode);
+ goto done;
+ }
+
+ __current->root = dnode;
+
+ unlock_dnode(dnode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(char*, getcwd, char*, buf, size_t, size)
+{
+ int errno = 0;
+ char* ret_ptr = 0;
+ if (size < 2) {
+ errno = ERANGE;
+ goto done;
+ }
+
+ size_t len = 0;
+
+ if (!__current->cwd) {
+ *buf = VFS_PATH_DELIM;
+ len = 1;
+ } else {
+ len = vfs_get_path(__current->cwd, buf, size, 0);
+ if (len == size) {
+ errno = ERANGE;
+ goto done;
+ }
+ }
+
+ buf[len] = '\0';
+
+ ret_ptr = buf;
+
+done:
+ syscall_result(errno);
+ return ret_ptr;
+}
+
+int
+vfs_do_rename(struct v_dnode* current, struct v_dnode* target)
+{
+ int errno = 0;
+ if (current->inode->id == target->inode->id) {
+ // hard link
+ return 0;
+ }
+
+ if ((errno = vfs_check_writable(current))) {
+ return errno;
+ }
+
+ if (current->ref_count > 1 || target->ref_count > 1) {
+ return EBUSY;
+ }
+
+ if (current->super_block != target->super_block) {
+ return EXDEV;
+ }
+
+ struct v_dnode* oldparent = current->parent;
+ struct v_dnode* newparent = target->parent;
+
+ lock_dnode(current);
+ lock_dnode(target);
+
+ if (oldparent)
+ lock_dnode(oldparent);
+ if (newparent)
+ lock_dnode(newparent);
+
+ if (!llist_empty(&target->children)) {
+ errno = ENOTEMPTY;
+ unlock_dnode(target);
+ goto cleanup;
+ }
+
+ if ((errno =
+ current->inode->ops->rename(current->inode, current, target))) {
+ unlock_dnode(target);
+ goto cleanup;
+ }
+
+ // re-position current
+ hstrcpy(¤t->name, &target->name);
+ vfs_dcache_rehash(newparent, current);
+
+ // detach target
+ vfs_d_free(target);
+
+ unlock_dnode(target);
+
+cleanup:
+ unlock_dnode(current);
+
+ if (oldparent)
+ unlock_dnode(oldparent);
+ if (newparent)
+ unlock_dnode(newparent);
+
+ return errno;
+}
+
+__DEFINE_LXSYSCALL2(int, rename, const char*, oldpath, const char*, newpath)
+{
+ struct v_dnode *cur, *target_parent, *target;
+ struct hstr name = HSTR(valloc(VFS_NAME_MAXLEN), 0);
+ int errno = 0;
+
+ if ((errno = vfs_walk_proc(oldpath, &cur, NULL, 0))) {
+ goto done;
+ }
+
+ if ((errno = vfs_walk(
+ __current->cwd, newpath, &target_parent, &name, VFS_WALK_PARENT))) {
+ goto done;
+ }
+
+ errno = vfs_walk(target_parent, name.value, &target, NULL, 0);
+ if (errno == ENOENT) {
+ target = vfs_d_alloc(target_parent, &name);
+ vfs_dcache_add(target_parent, target);
+ } else if (errno) {
+ goto done;
+ }
+
+ if (!target) {
+ errno = ENOMEM;
+ goto done;
+ }
+
+ errno = vfs_do_rename(cur, target);
+
+done:
+ vfree((void*)name.value);
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL2(int, fstat, int, fd, struct file_stat*, stat)
+{
+ int errno = 0;
+ struct v_fd* fds;
+
+ if ((errno = vfs_getfd(fd, &fds))) {
+ goto done;
+ }
+
+ struct v_inode* vino = fds->file->inode;
+ struct device* fdev = vino->sb->dev;
+
+ stat->st_ino = vino->id;
+ stat->st_blocks = vino->lb_usage;
+ stat->st_size = vino->fsize;
+ stat->st_blksize = vino->sb->blksize;
+ stat->st_nlink = vino->link_count;
+ stat->st_uid = vino->uid;
+ stat->st_gid = vino->gid;
+
+ stat->st_ctim = vino->ctime;
+ stat->st_atim = vino->atime;
+ stat->st_mtim = vino->mtime;
+
+ stat->st_mode = (vino->itype << 16) | vino->acl;
+
+ stat->st_ioblksize = PAGE_SIZE;
+
+ if (check_device_node(vino)) {
+ struct device* rdev = resolve_device(vino->data);
+ if (!rdev) {
+ errno = EINVAL;
+ goto done;
+ }
+
+ stat->st_rdev = (dev_t){.meta = rdev->ident.fn_grp,
+ .unique = rdev->ident.unique,
+ .index = dev_uid(rdev) };
+ }
+
+ if (fdev) {
+ stat->st_dev = (dev_t){.meta = fdev->ident.fn_grp,
+ .unique = fdev->ident.unique,
+ .index = dev_uid(fdev) };
+ }
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL4(int, fchmodat, int, fd,
+ const char*, path, int, mode, int, flags)
+{
+ int errno;
+ struct v_dnode *dnode;
+ struct v_inode* inode;
+
+ errno = vfs_walkat(fd, path, flags, &dnode);
+ if (errno) {
+ goto done;
+ }
+
+ errno = vfs_check_writable(dnode);
+ if (errno) {
+ return errno;
+ }
+
+ inode = dnode->inode;
+ lock_inode(inode);
+
+ if (!current_is_root()) {
+ mode = mode & FSACL_RWXMASK;
+ }
+
+ inode->acl = mode;
+ __vfs_touch_inode(inode, INODE_MODIFY);
+
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL5(int, fchownat, int, fd,
+ const char*, path, uid_t, uid, gid_t, gid, int, flags)
+{
+ int errno;
+ struct v_dnode *dnode;
+ struct v_inode *inode;
+
+ errno = vfs_walkat(fd, path, flags, &dnode);
+ if (errno) {
+ goto done;
+ }
+
+ errno = vfs_check_writable(dnode);
+ if (errno) {
+ return errno;
+ }
+
+ inode = dnode->inode;
+ lock_inode(inode);
+
+ inode->uid = uid;
+ inode->gid = gid;
+ __vfs_touch_inode(inode, INODE_MODIFY);
+
+ unlock_inode(inode);
+
+done:
+ return DO_STATUS(errno);
+}
+
+__DEFINE_LXSYSCALL4(int, faccessat, int, fd,
+ const char*, path, int, amode, int, flags)
+{
+ int errno, acl;
+ struct v_dnode *dnode;
+ struct v_inode *inode;
+ struct user_scope* uscope;
+
+ uid_t tuid;
+ gid_t tgid;
+
+ errno = vfs_walkat(fd, path, flags, &dnode);
+ if (errno) {
+ goto done;
+ }
+
+ if ((flags & AT_EACCESS)) {
+ tuid = current_euid();
+ tgid = current_egid();
+ }
+ else {
+ uscope = current_user_scope();
+ tuid = uscope->ruid;
+ tgid = uscope->rgid;
+ }
+
+ inode = dnode->inode;
+
+ acl = inode->acl;
+ acl &= amode;
+ acl &= check_acl_between(inode->uid, inode->gid, tuid, tgid);
+ if (!acl) {
+ errno = EACCESS;
+ }
+
+done:
+ return DO_STATUS(errno);
git://scm.lunaixsky.com / lunaix-os.git / blobdiff