summaryrefslogtreecommitdiffstats
path: root/cpukit/libfs/src/dosfs/msdos_dir.c
diff options
context:
space:
mode:
Diffstat (limited to 'cpukit/libfs/src/dosfs/msdos_dir.c')
-rw-r--r--cpukit/libfs/src/dosfs/msdos_dir.c706
1 files changed, 706 insertions, 0 deletions
diff --git a/cpukit/libfs/src/dosfs/msdos_dir.c b/cpukit/libfs/src/dosfs/msdos_dir.c
new file mode 100644
index 0000000000..7e6bcf3db6
--- /dev/null
+++ b/cpukit/libfs/src/dosfs/msdos_dir.c
@@ -0,0 +1,706 @@
+/*
+ * MSDOS directory handlers implementation
+ *
+ * Copyright (C) 2001 OKTET Ltd., St.-Petersburg, Russia
+ * Author: Eugeny S. Mints <Eugeny.Mints@oktet.ru>
+ *
+ * The license and distribution terms for this file may be
+ * found in the file LICENSE in this distribution or at
+ * http://www.rtems.com/license/LICENSE.
+ *
+ * @(#) $Id$
+ */
+#if HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <ctype.h>
+#include <stdlib.h>
+#include <unistd.h>
+#include <errno.h>
+#include <rtems/libio_.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+
+#include <dirent.h>
+
+#include "fat.h"
+#include "fat_fat_operations.h"
+#include "fat_file.h"
+
+#include "msdos.h"
+
+/* msdos_dir_open --
+ * Open fat-file which correspondes to the directory being opened and
+ * set offset field of file control block to zero.
+ *
+ * PARAMETERS:
+ * iop - file control block
+ * pathname - name
+ * flag - flags
+ * mode - mode
+ *
+ * RETURNS:
+ * RC_OK, if directory opened successfully, or -1 if error occured (errno
+ * set apropriately)
+ */
+int
+msdos_dir_open(rtems_libio_t *iop, const char *pathname, uint32_t flag,
+ uint32_t mode)
+{
+ int rc = RC_OK;
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ msdos_fs_info_t *fs_info = iop->pathinfo.mt_entry->fs_info;
+ fat_file_fd_t *fat_fd = iop->pathinfo.node_access;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one( EIO );
+
+ rc = fat_file_reopen(fat_fd);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ iop->offset = 0;
+ rtems_semaphore_release(fs_info->vol_sema);
+ return RC_OK;
+}
+
+/* msdos_dir_close --
+ * Close fat-file which correspondes to the directory being closed
+ *
+ * PARAMETERS:
+ * iop - file control block
+ *
+ * RETURNS:
+ * RC_OK, if directory closed successfully, or -1 if error occured (errno
+ * set apropriately.
+ */
+int
+msdos_dir_close(rtems_libio_t *iop)
+{
+ int rc = RC_OK;
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ msdos_fs_info_t *fs_info = iop->pathinfo.mt_entry->fs_info;
+ fat_file_fd_t *fat_fd = iop->pathinfo.node_access;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one( EIO );
+
+ rc = fat_file_close(iop->pathinfo.mt_entry, fat_fd);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ rtems_semaphore_release(fs_info->vol_sema);
+ return RC_OK;
+}
+
+/* msdos_format_dirent_with_dot --
+ * This routine convert a (short) MSDOS filename as present on disk
+ * (fixed 8+3 characters, filled with blanks, without separator dot)
+ * to a "normal" format, with between 0 and 8 name chars,
+ * a separating dot and up to 3 extension characters
+ * Rules to work:
+ * - copy any (0-8) "name" part characters that are non-blank
+ * - if an extension exists, append a dot
+ * - copy any (0-3) non-blank extension characters
+ * - append a '\0' (dont count it for the rturn code
+ *
+ * PARAMETERS:
+ * dst: pointer to destination char array (must be big enough)
+ * src: pointer to source characters
+ *
+ *
+ * RETURNS:
+ * the number of bytes (without trailing '\0'(written to destination
+ */
+static ssize_t
+msdos_format_dirent_with_dot(char *dst,const char *src)
+{
+ ssize_t len;
+ int i;
+ const char *src_tmp;
+
+ /*
+ * find last non-blank character of base name
+ */
+ for ((i = MSDOS_SHORT_BASE_LEN ,
+ src_tmp = src + MSDOS_SHORT_BASE_LEN-1);
+ ((i > 0) &&
+ (*src_tmp == ' '));
+ i--,src_tmp--)
+ {};
+ /*
+ * copy base name to destination
+ */
+ src_tmp = src;
+ len = i;
+ while (i-- > 0) {
+ *dst++ = tolower((unsigned char)(*src_tmp++));
+ }
+ /*
+ * find last non-blank character of extension
+ */
+ for ((i = MSDOS_SHORT_EXT_LEN ,
+ src_tmp = src + MSDOS_SHORT_BASE_LEN+MSDOS_SHORT_EXT_LEN-1);
+ ((i > 0) &&
+ (*src_tmp == ' '));
+ i--,src_tmp--)
+ {};
+ /*
+ * extension is not empty
+ */
+ if (i > 0) {
+ *dst++ = '.'; /* append dot */
+ len += i + 1; /* extension + dot */
+ src_tmp = src + MSDOS_SHORT_BASE_LEN;
+ while (i-- > 0) {
+ *dst++ = tolower((unsigned char)(*src_tmp++));
+ len++;
+ }
+ }
+ *dst = '\0'; /* terminate string */
+
+ return len;
+}
+
+/* msdos_dir_read --
+ * This routine will read the next directory entry based on the directory
+ * offset. The offset should be equal to -n- time the size of an
+ * individual dirent structure. If n is not an integer multiple of the
+ * sizeof a dirent structure, an integer division will be performed to
+ * determine directory entry that will be returned in the buffer. Count
+ * should reflect -m- times the sizeof dirent bytes to be placed in the
+ * buffer.
+ * If there are not -m- dirent elements from the current directory
+ * position to the end of the exisiting file, the remaining entries will
+ * be placed in the buffer and the returned value will be equal to
+ * -m actual- times the size of a directory entry.
+ *
+ * PARAMETERS:
+ * iop - file control block
+ * buffer - buffer provided by user
+ * count - count of bytes to read
+ *
+ * RETURNS:
+ * the number of bytes read on success, or -1 if error occured (errno
+ * set apropriately).
+ */
+ssize_t
+msdos_dir_read(rtems_libio_t *iop, void *buffer, size_t count)
+{
+ int rc = RC_OK;
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ msdos_fs_info_t *fs_info = iop->pathinfo.mt_entry->fs_info;
+ fat_file_fd_t *fat_fd = iop->pathinfo.node_access;
+ fat_file_fd_t *tmp_fat_fd = NULL;
+ struct dirent tmp_dirent;
+ uint32_t start = 0;
+ ssize_t ret = 0;
+ uint32_t cmpltd = 0;
+ uint32_t j = 0, i = 0;
+ uint32_t bts2rd = 0;
+ uint32_t cur_cln = 0;
+ uint32_t lfn_start = FAT_FILE_SHORT_NAME;
+ uint8_t lfn_checksum = 0;
+ int lfn_entries = 0;
+
+ /*
+ * cast start and count - protect against using sizes that are not exact
+ * multiples of the -dirent- size. These could result in unexpected
+ * results
+ */
+ start = iop->offset / sizeof(struct dirent);
+ count = (count / sizeof(struct dirent)) * sizeof(struct dirent);
+
+ /*
+ * optimization: we know that root directory for FAT12/16 volumes is
+ * sequential set of sectors and any cluster is sequential set of sectors
+ * too, so read such set of sectors is quick operation for low-level IO
+ * layer.
+ */
+ bts2rd = (FAT_FD_OF_ROOT_DIR(fat_fd) &&
+ (fs_info->fat.vol.type & (FAT_FAT12 | FAT_FAT16))) ?
+ fat_fd->fat_file_size :
+ fs_info->fat.vol.bpc;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one(EIO);
+
+ while (count > 0)
+ {
+ /*
+ * fat-file is already opened by open call, so read it
+ * Always read directory fat-file from the beggining because of MSDOS
+ * directories feature :( - we should count elements currently
+ * present in the directory because there may be holes :)
+ */
+ ret = fat_file_read(iop->pathinfo.mt_entry, fat_fd, (j * bts2rd),
+ bts2rd, fs_info->cl_buf);
+ if (ret < MSDOS_DIRECTORY_ENTRY_STRUCT_SIZE)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ rtems_set_errno_and_return_minus_one(EIO);
+ }
+
+ for (i = 0; i < ret; i += MSDOS_DIRECTORY_ENTRY_STRUCT_SIZE)
+ {
+ char* entry = (char*) fs_info->cl_buf + i;
+
+ /*
+ * Is this directory from here on empty ?
+ */
+ if ((*MSDOS_DIR_ENTRY_TYPE(entry)) ==
+ MSDOS_THIS_DIR_ENTRY_AND_REST_EMPTY)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return cmpltd;
+ }
+
+ /* Is the directory entry empty */
+ if ((*MSDOS_DIR_ENTRY_TYPE(entry)) == MSDOS_THIS_DIR_ENTRY_EMPTY)
+ continue;
+
+ /* Is the directory entry empty a volume label */
+ if (((*MSDOS_DIR_ATTR(entry)) & MSDOS_ATTR_VOLUME_ID) &&
+ ((*MSDOS_DIR_ATTR(entry) & MSDOS_ATTR_LFN_MASK) != MSDOS_ATTR_LFN))
+ continue;
+
+ /*
+ * Check the attribute to see if the entry is for a long file
+ * name.
+ */
+ if ((*MSDOS_DIR_ATTR(entry) & MSDOS_ATTR_LFN_MASK) ==
+ MSDOS_ATTR_LFN)
+ {
+ int o;
+ char* p;
+ int q;
+
+ /*
+ * Is this is the first entry of a LFN ?
+ */
+ if (lfn_start == FAT_FILE_SHORT_NAME)
+ {
+ /*
+ * The first entry must have the last long entry flag set.
+ */
+ if ((*MSDOS_DIR_ENTRY_TYPE(entry) &
+ MSDOS_LAST_LONG_ENTRY) == 0)
+ continue;
+
+ /*
+ * Remember the start location of the long file name.
+ */
+ lfn_start =
+ ((j * bts2rd) + i) / MSDOS_DIRECTORY_ENTRY_STRUCT_SIZE;
+
+ /*
+ * Get the number of entries so we can count down and
+ * also the checksum of the short entry.
+ */
+ lfn_entries = (*MSDOS_DIR_ENTRY_TYPE(entry) &
+ MSDOS_LAST_LONG_ENTRY_MASK);
+ lfn_checksum = *MSDOS_DIR_LFN_CHECKSUM(entry);
+ memset (tmp_dirent.d_name, 0, sizeof(tmp_dirent.d_name));
+ }
+
+ /*
+ * If the entry number or the check sum do not match
+ * forget this series of long directory entries. These could
+ * be orphaned entries depending on the history of the
+ * disk.
+ */
+ if ((lfn_entries != (*MSDOS_DIR_ENTRY_TYPE(entry) &
+ MSDOS_LAST_LONG_ENTRY_MASK)) ||
+ (lfn_checksum != *MSDOS_DIR_LFN_CHECKSUM(entry)))
+ {
+ lfn_start = FAT_FILE_SHORT_NAME;
+ continue;
+ }
+
+ /*
+ * Extract the file name into the directory entry. The data is
+ * stored in UNICODE characters (16bit). No translation is
+ * currently supported.
+ *
+ * The DOS maximum length is 255 characters without the
+ * trailing nul character. We need to range check the length to
+ * fit in the directory entry name field.
+ */
+
+ lfn_entries--;
+ p = entry + 1;
+ o = lfn_entries * MSDOS_LFN_LEN_PER_ENTRY;
+
+ for (q = 0; q < MSDOS_LFN_LEN_PER_ENTRY; q++)
+ {
+ if (o >= (sizeof(tmp_dirent.d_name) - 1))
+ break;
+
+ tmp_dirent.d_name[o++] = *p;
+
+ if (*p == '\0')
+ break;
+
+ switch (q)
+ {
+ case 4:
+ p += 5;
+ break;
+ case 10:
+ p += 4;
+ break;
+ default:
+ p += 2;
+ break;
+ }
+ }
+ }
+ else
+ {
+ fat_dir_pos_t dir_pos;
+
+ /*
+ * Skip active entries until get the entry to start from.
+ */
+ if (start)
+ {
+ lfn_start = FAT_FILE_SHORT_NAME;
+ start--;
+ continue;
+ }
+
+ /*
+ * Move the entry to the return buffer
+ *
+ * unfortunately there is no method to extract ino except to
+ * open fat-file descriptor :( ... so, open it
+ */
+
+ /* get number of cluster we are working with */
+ rc = fat_file_ioctl(iop->pathinfo.mt_entry, fat_fd, F_CLU_NUM,
+ j * bts2rd, &cur_cln);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ fat_dir_pos_init(&dir_pos);
+ dir_pos.sname.cln = cur_cln;
+ dir_pos.sname.ofs = i;
+ rc = fat_file_open(iop->pathinfo.mt_entry, &dir_pos, &tmp_fat_fd);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ /* fill in dirent structure */
+ /* XXX: from what and in what d_off should be computed ?! */
+ tmp_dirent.d_off = start + cmpltd;
+ tmp_dirent.d_reclen = sizeof(struct dirent);
+ tmp_dirent.d_ino = tmp_fat_fd->ino;
+
+ /*
+ * If a long file name check if the correct number of
+ * entries have been found and if the checksum is correct.
+ * If not return the short file name.
+ */
+ if (lfn_start != FAT_FILE_SHORT_NAME)
+ {
+ uint8_t cs = 0;
+ uint8_t* p = (uint8_t*) entry;
+ int i;
+
+ for (i = 0; i < 11; i++, p++)
+ cs = ((cs & 1) ? 0x80 : 0) + (cs >> 1) + *p;
+
+ if (lfn_entries || (lfn_checksum != cs))
+ lfn_start = FAT_FILE_SHORT_NAME;
+ }
+
+ if (lfn_start == FAT_FILE_SHORT_NAME)
+ {
+ /*
+ * convert dir entry from fixed 8+3 format (without dot)
+ * to 0..8 + 1dot + 0..3 format
+ */
+ tmp_dirent.d_namlen = msdos_format_dirent_with_dot(
+ tmp_dirent.d_name, entry); /* src text */
+ }
+ else
+ {
+ tmp_dirent.d_namlen = strlen(tmp_dirent.d_name);
+ }
+
+ memcpy(buffer + cmpltd, &tmp_dirent, sizeof(struct dirent));
+
+ iop->offset = iop->offset + sizeof(struct dirent);
+ cmpltd += (sizeof(struct dirent));
+ count -= (sizeof(struct dirent));
+
+ /* inode number extracted, close fat-file */
+ rc = fat_file_close(iop->pathinfo.mt_entry, tmp_fat_fd);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+ }
+
+ if (count <= 0)
+ break;
+ }
+ j++;
+ }
+
+ rtems_semaphore_release(fs_info->vol_sema);
+ return cmpltd;
+}
+
+/* msdos_dir_write --
+ * no write for directory
+ */
+
+/* msdos_dir_lseek --
+ *
+ * This routine will behave in one of three ways based on the state of
+ * argument whence. Based on the state of its value the offset argument will
+ * be interpreted using one of the following methods:
+ *
+ * SEEK_SET - offset is the absolute byte offset from the start of the
+ * logical start of the dirent sequence that represents the
+ * directory
+ * SEEK_CUR - offset is used as the relative byte offset from the current
+ * directory position index held in the iop structure
+ * SEEK_END - N/A --> This will cause an assert.
+ *
+ * PARAMETERS:
+ * iop - file control block
+ * offset - offset
+ * whence - predefine directive
+ *
+ * RETURNS:
+ * RC_OK on success, or -1 if error occured (errno
+ * set apropriately).
+ */
+rtems_off64_t
+msdos_dir_lseek(rtems_libio_t *iop, rtems_off64_t offset, int whence)
+{
+ switch (whence)
+ {
+ case SEEK_SET:
+ case SEEK_CUR:
+ break;
+ /*
+ * Movement past the end of the directory via lseek is not a
+ * permitted operation
+ */
+ case SEEK_END:
+ default:
+ rtems_set_errno_and_return_minus_one( EINVAL );
+ break;
+ }
+ return RC_OK;
+}
+
+/* msdos_dir_stat --
+ *
+ * This routine will obtain the following information concerning the current
+ * directory:
+ * st_dev device id
+ * st_ino node serial number :)
+ * st_mode mode extracted from the node
+ * st_size total size in bytes
+ * st_blksize blocksize for filesystem I/O
+ * st_blocks number of blocks allocated
+ * stat_mtime time of last modification
+ *
+ * PARAMETERS:
+ * loc - this directory
+ * buf - stat buffer provided by user
+ *
+ * RETURNS:
+ * RC_OK and filled stat buffer on success, or -1 if error occured (errno
+ * set apropriately).
+ */
+int
+msdos_dir_stat(
+ rtems_filesystem_location_info_t *loc,
+ struct stat *buf
+ )
+{
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ msdos_fs_info_t *fs_info = loc->mt_entry->fs_info;
+ fat_file_fd_t *fat_fd = loc->node_access;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one(EIO);
+
+ buf->st_dev = fs_info->fat.vol.dev;
+ buf->st_ino = fat_fd->ino;
+ buf->st_mode = S_IFDIR | S_IRWXU | S_IRWXG | S_IRWXO;
+ buf->st_rdev = 0ll;
+ buf->st_size = fat_fd->fat_file_size;
+ buf->st_blocks = fat_fd->fat_file_size >> FAT_SECTOR512_BITS;
+ buf->st_blksize = fs_info->fat.vol.bps;
+ buf->st_mtime = fat_fd->mtime;
+
+ rtems_semaphore_release(fs_info->vol_sema);
+ return RC_OK;
+}
+
+/* msdos_dir_truncate --
+ * No truncate for directory.
+ *
+ * PARAMETERS:
+ *
+ * RETURNS:
+ *
+ */
+
+/* msdos_dir_sync --
+ * The following routine does a syncronization on a MSDOS directory node.
+ * DIR_WrtTime, DIR_WrtDate and DIR_fileSize fields of 32 Bytes Directory
+ * Entry Structure should not be updated for directories, so only call
+ * to corresponding fat-file routine.
+ *
+ * PARAMETERS:
+ * iop - file control block
+ *
+ * RETURNS:
+ * RC_OK on success, or -1 if error occured (errno set apropriately).
+ */
+int
+msdos_dir_sync(rtems_libio_t *iop)
+{
+ int rc = RC_OK;
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ fat_file_fd_t *fat_fd = iop->pathinfo.node_access;
+ msdos_fs_info_t *fs_info = iop->pathinfo.mt_entry->fs_info;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one(EIO);
+
+ rc = fat_file_datasync(iop->pathinfo.mt_entry, fat_fd);
+
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+}
+
+/* msdos_dir_chmod --
+ * Change the attributes of the directory. This currently does
+ * nothing and returns no error.
+ *
+ * PARAMETERS:
+ * pathloc - node description
+ * mode - the new mode
+ *
+ * RETURNS:
+ * RC_OK always
+ */
+int
+msdos_dir_chmod(rtems_filesystem_location_info_t *pathloc,
+ mode_t mode)
+{
+ return RC_OK;
+}
+
+/* msdos_dir_rmnod --
+ * Remove directory node.
+ *
+ * Check that this directory node is not opened as fat-file, is empty and
+ * not filesystem root node. If all this conditions met then delete.
+ *
+ * PARAMETERS:
+ * pathloc - node description
+ *
+ * RETURNS:
+ * RC_OK on success, or -1 if error occured (errno set apropriately).
+ */
+int
+msdos_dir_rmnod(rtems_filesystem_location_info_t *parent_pathloc,
+ rtems_filesystem_location_info_t *pathloc)
+{
+ int rc = RC_OK;
+ rtems_status_code sc = RTEMS_SUCCESSFUL;
+ msdos_fs_info_t *fs_info = pathloc->mt_entry->fs_info;
+ fat_file_fd_t *fat_fd = pathloc->node_access;
+ bool is_empty = false;
+
+ sc = rtems_semaphore_obtain(fs_info->vol_sema, RTEMS_WAIT,
+ MSDOS_VOLUME_SEMAPHORE_TIMEOUT);
+ if (sc != RTEMS_SUCCESSFUL)
+ rtems_set_errno_and_return_minus_one(EIO);
+
+ /*
+ * We deny attempts to delete open directory (if directory is current
+ * directory we assume it is open one)
+ */
+ if (fat_fd->links_num > 1)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ rtems_set_errno_and_return_minus_one(EBUSY);
+ }
+
+ /*
+ * You cannot remove a node that still has children
+ */
+ rc = msdos_dir_is_empty(pathloc->mt_entry, fat_fd, &is_empty);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ if (!is_empty)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ rtems_set_errno_and_return_minus_one(ENOTEMPTY);
+ }
+
+ /*
+ * You cannot remove the file system root node.
+ */
+ if (pathloc->mt_entry->mt_fs_root.node_access == pathloc->node_access)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ rtems_set_errno_and_return_minus_one(EBUSY);
+ }
+
+ /*
+ * You cannot remove a mountpoint.
+ * not used - mount() not implemenetd yet.
+ */
+
+ /* mark file removed */
+ rc = msdos_set_first_char4file_name(pathloc->mt_entry, &fat_fd->dir_pos,
+ MSDOS_THIS_DIR_ENTRY_EMPTY);
+ if (rc != RC_OK)
+ {
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+ }
+
+ fat_file_mark_removed(pathloc->mt_entry, fat_fd);
+
+ rtems_semaphore_release(fs_info->vol_sema);
+ return rc;
+}