Filesystem: Reference counting for locations

 o A new data structure rtems_filesystem_global_location_t was
   introduced to be used for
    o the mount point location in the mount table entry,
    o the file system root location in the mount table entry,
    o the root directory location in the user environment, and
    o the current directory location in the user environment.
   During the path evaluation global start locations are obtained to
   ensure that the current file system instance will be not unmounted in
   the meantime.
 o The user environment uses now reference counting and is protected
   from concurrent access.
 o The path evaluation process was completely rewritten and simplified.
   The IMFS, RFS, NFS, and DOSFS use now a generic path evaluation
   method.  Recursive calls in the path evaluation have been replaced
   with iteration to avoid stack overflows.  Only the evaluation of
   symbolic links is recursive.  No dynamic memory allocations and
   intermediate buffers are used in the high level path evaluation.  No
   global locks are held during the file system instance specific path
   evaluation process.
 o Recursive symbolic link evaluation is now limited by
   RTEMS_FILESYSTEM_SYMLOOP_MAX.  Applications can retrieve this value
   via sysconf().
 o The device file system (devFS) uses now no global variables and
   allocation from the workspace.  Node names are allocated from the
   heap.
 o The upper layer lseek() performs now some parameter checks.
 o The upper layer ftruncate() performs now some parameter checks.
 o unmask() is now restricted to the RWX flags and protected from
   concurrent access.
 o The fchmod_h and rmnod_h file system node handlers are now a file
   system operation.
 o The unlink_h operation has been removed.  All nodes are now destroyed
   with the rmnod_h operation.
 o New lock_h, unlock_h, clonenod_h, and are_nodes_equal_h file system
   operations.
 o The path evaluation and file system operations are now protected by
   per file system instance lock and unlock operations.
 o Fix and test file descriptor duplicate in fcntl().
 o New test fstests/fsnofs01.

1 files changed, 144 insertions, 182 deletions

diff --git a/cpukit/libcsupport/src/mount.c b/cpukit/libcsupport/src/mount.c
index 00a07aab3a..47c08f7dba 100644
--- a/cpukit/libcsupport/src/mount.c
+++ b/cpukit/libcsupport/src/mount.c
@@ -20,22 +20,15 @@
  */
 
 #if HAVE_CONFIG_H
-#include "config.h"
+  #include "config.h"
 #endif
 
-#include <sys/types.h>
-#include <sys/stat.h>
-#include <rtems/chain.h>
-#include <rtems/seterr.h>
-#include <fcntl.h>
-#include <unistd.h>
-#include <errno.h>
 #include <stdlib.h>
 #include <string.h>
 
 #include <rtems/libio_.h>
 
-static RTEMS_CHAIN_DEFINE_EMPTY(mount_chain);
+RTEMS_CHAIN_DEFINE_EMPTY(rtems_filesystem_mount_table);
 
 /*
  * Default pathconfs.
@@ -55,14 +48,6 @@ const rtems_filesystem_limits_and_options_t rtems_filesystem_default_pathconf =
    0     /* posix_vdisable: special char processing, 0=no, 1=yes */
 };
 
-static bool is_node_fs_root(
-  const rtems_filesystem_mount_table_entry_t *mt_entry,
-  void *arg
-)
-{
-  return arg == mt_entry->mt_fs_root.node_access;
-}
-
 static rtems_filesystem_mount_table_entry_t *alloc_mount_table_entry(
   const char *source_or_null,
   const char *target_or_null,
@@ -76,11 +61,15 @@ static rtems_filesystem_mount_table_entry_t *alloc_mount_table_entry(
     strlen( source_or_null ) + 1 : 0;
   size_t target_size = strlen( target ) + 1;
   size_t size = sizeof( rtems_filesystem_mount_table_entry_t )
-    + filesystemtype_size + source_size + target_size;
+    + filesystemtype_size + source_size + target_size
+    + sizeof( rtems_filesystem_global_location_t );
   rtems_filesystem_mount_table_entry_t *mt_entry = calloc( 1, size );
 
   if ( mt_entry != NULL ) {
-    char *str = (char *) mt_entry + sizeof( *mt_entry );
+    rtems_filesystem_global_location_t *mt_fs_root =
+      (rtems_filesystem_global_location_t *)
+        ((char *) mt_entry + sizeof( *mt_entry ));
+    char *str = (char *) mt_fs_root + sizeof( *mt_fs_root );
 
     memcpy( str, filesystemtype, filesystemtype_size );
     mt_entry->type = str;
@@ -92,6 +81,23 @@ static rtems_filesystem_mount_table_entry_t *alloc_mount_table_entry(
 
     memcpy( str, target, target_size );
     mt_entry->target = str;
+    str += target_size;
+
+    mt_entry->mounted = true;
+    mt_entry->mt_fs_root = mt_fs_root;
+    mt_entry->pathconf_limits_and_options = rtems_filesystem_default_pathconf;
+
+    mt_fs_root->location.mt_entry = mt_entry;
+    mt_fs_root->reference_count = 1;
+
+    rtems_chain_initialize(
+      &mt_entry->location_chain,
+      mt_fs_root,
+      1,
+      sizeof(*mt_fs_root)
+    );
+  } else {
+    free( mt_entry );
   }
 
   *target_length_ptr = target_size - 1;
@@ -99,184 +105,140 @@ static rtems_filesystem_mount_table_entry_t *alloc_mount_table_entry(
   return mt_entry;
 }
 
-/*
- *  mount
- *
- *  This routine will attempt to mount a new file system at the specified
- *  mount point. A series of tests will be run to determine if any of the
- *  following reasons exist to prevent the mount operation:
- *
- * 	1) The file system type or options are not valid
- *	2) No new file system root node is specified
- * 	3) The selected file system has already been mounted
- * 	4) The mount point exists with the proper permissions to allow mounting
- *	5) The selected mount point already has a file system mounted to it
- *
- */
-
-int mount(
-  const char                 *source,
-  const char                 *target,
-  const char                 *filesystemtype,
-  rtems_filesystem_options_t options,
-  const void                 *data
+static int register_subordinate_file_system(
+  rtems_filesystem_mount_table_entry_t *mt_entry,
+  const char *target
 )
 {
-  rtems_filesystem_fsmount_me_t mount_h = NULL;
-  rtems_filesystem_location_info_t      loc;
-  rtems_filesystem_mount_table_entry_t *mt_entry = NULL;
-  rtems_filesystem_location_info_t     *loc_to_free = NULL;
-  bool has_target = target != NULL;
-  size_t target_length = 0;
-
-  /*
-   *  Are the file system options valid?
-   */
-
-  if ( options != RTEMS_FILESYSTEM_READ_ONLY &&
-       options != RTEMS_FILESYSTEM_READ_WRITE )
-    rtems_set_errno_and_return_minus_one( EINVAL );
-
-  /*
-   *  Get mount handler
-   */
-  mount_h = rtems_filesystem_get_mount_handler( filesystemtype );
-  if ( !mount_h )
-    rtems_set_errno_and_return_minus_one( EINVAL );
-
-  /*
-   * Allocate a mount table entry
-   */
-  mt_entry = alloc_mount_table_entry(
-    source,
-    target,
-    filesystemtype,
-    &target_length
-  );
-  if ( !mt_entry )
-    rtems_set_errno_and_return_minus_one( ENOMEM );
-
-  mt_entry->mt_fs_root.mt_entry = mt_entry;
-  mt_entry->options = options;
-  mt_entry->pathconf_limits_and_options = rtems_filesystem_default_pathconf;
-
-  /*
-   *  The mount_point should be a directory with read/write/execute
-   *  permissions in the existing tree.
-   */
-
-  if ( has_target ) {
-    if ( rtems_filesystem_evaluate_path(
-           target, target_length, RTEMS_LIBIO_PERMS_RWX, &loc, true ) == -1 )
-      goto cleanup_and_bail;
-
-    loc_to_free = &loc;
-
-    /*
-     *  Test to see if it is a directory
-     */
-
-    if ( loc.ops->node_type_h( &loc ) != RTEMS_FILESYSTEM_DIRECTORY ) {
-      errno = ENOTDIR;
-      goto cleanup_and_bail;
+  int rv = 0;
+  rtems_filesystem_eval_path_context_t ctx;
+  int eval_flags = RTEMS_LIBIO_PERMS_RWX
+    | RTEMS_LIBIO_FOLLOW_LINK;
+  rtems_filesystem_location_info_t *currentloc =
+    rtems_filesystem_eval_path_start( &ctx, target, eval_flags );
+
+  if ( !rtems_filesystem_location_is_root( currentloc ) ) {
+    rtems_filesystem_location_info_t targetloc;
+    rtems_filesystem_global_location_t *mt_point_node;
+
+    rtems_filesystem_eval_path_extract_currentloc( &ctx, &targetloc );
+    mt_point_node = rtems_filesystem_location_transform_to_global( &targetloc );
+    mt_entry->mt_point_node = mt_point_node;
+    rv = (*mt_point_node->location.ops->mount_h)( mt_entry );
+    if ( rv == 0 ) {
+      rtems_filesystem_mt_lock();
+      rtems_chain_append_unprotected(
+        &rtems_filesystem_mount_table,
+        &mt_entry->mt_node
+      );
+      rtems_filesystem_mt_unlock();
+    } else {
+      rtems_filesystem_global_location_release( mt_point_node );
     }
+  } else {
+    rtems_filesystem_eval_path_error( &ctx, EBUSY );
+    rv = -1;
+  }
 
-    /*
-     *  You can only mount one file system onto a single mount point.
-     */
+  rtems_filesystem_eval_path_cleanup( &ctx );
 
-    if ( rtems_filesystem_mount_iterate( is_node_fs_root, loc.node_access ) ) {
-      errno = EBUSY;
-      goto cleanup_and_bail;
-    }
+  return rv;
+}
 
-    /*
-     *  This must be a good mount point, so move the location information
-     *  into the allocated mount entry.  Note:  the information that
-     *  may have been allocated in loc should not be sent to freenode
-     *  until the system is unmounted.  It may be needed to correctly
-     *  traverse the tree.
-     */
-
-    mt_entry->mt_point_node.node_access = loc.node_access;
-    mt_entry->mt_point_node.handlers = loc.handlers;
-    mt_entry->mt_point_node.ops = loc.ops;
-    mt_entry->mt_point_node.mt_entry = loc.mt_entry;
-
-    /*
-     *  This link to the parent is only done when we are dealing with system
-     *  below the base file system
-     */
-
-    if ( loc.ops->mount_h( mt_entry ) ) {
-      goto cleanup_and_bail;
-    }
+static int register_root_file_system(
+  rtems_filesystem_mount_table_entry_t *mt_entry
+)
+{
+  int rv = 0;
+
+  rtems_filesystem_mt_lock();
+  if ( rtems_chain_is_empty( &rtems_filesystem_mount_table ) ) {
+    rtems_chain_append_unprotected(
+      &rtems_filesystem_mount_table,
+      &mt_entry->mt_node
+    );
   } else {
-    /*
-     * Do we already have a base file system ?
-     */
-    if ( !rtems_chain_is_empty( &mount_chain ) ) {
-      errno = EINVAL;
-      goto cleanup_and_bail;
-    }
-
-    /*
-     *  This is a mount of the base file system --> The
-     *  mt_point_node.node_access will be left to null to indicate that this
-     *  is the root of the entire file system.
-     */
+    errno = EINVAL;
+    rv = -1;
   }
-
-  if ( (*mount_h)( mt_entry, data ) ) {
-    /*
-     * Try to undo the mount operation
-     */
-    loc.ops->unmount_h( mt_entry );
-    goto cleanup_and_bail;
+  rtems_filesystem_mt_unlock();
+
+  if ( rv == 0 ) {
+    rtems_filesystem_global_location_t *new_fs_root =
+      rtems_filesystem_global_location_obtain( &mt_entry->mt_fs_root );
+    rtems_filesystem_global_location_t *new_fs_current =
+      rtems_filesystem_global_location_obtain( &mt_entry->mt_fs_root );
+
+    rtems_filesystem_global_location_assign(
+      &rtems_filesystem_root,
+      new_fs_root
+    );
+    rtems_filesystem_global_location_assign(
+      &rtems_filesystem_current,
+      new_fs_current
+    );
   }
 
-  /*
-   *  Add the mount table entry to the mount table chain
-   */
-  rtems_libio_lock();
-  rtems_chain_append( &mount_chain, &mt_entry->Node );
-  rtems_libio_unlock();
-
-  if ( !has_target )
-    rtems_filesystem_root = mt_entry->mt_fs_root;
-
-  return 0;
-
-cleanup_and_bail:
-
-  free( mt_entry );
-
-  if ( loc_to_free )
-    rtems_filesystem_freenode( loc_to_free );
-
-  return -1;
+  return rv;
 }
 
-bool rtems_filesystem_mount_iterate(
-  rtems_per_filesystem_mount_routine routine,
-  void *routine_arg
+int mount(
+  const char                 *source,
+  const char                 *target,
+  const char                 *filesystemtype,
+  rtems_filesystem_options_t options,
+  const void                 *data
 )
 {
-  rtems_chain_node *node = NULL;
-  bool stop = false;
-
-  rtems_libio_lock();
-  for (
-    node = rtems_chain_first( &mount_chain );
-    !rtems_chain_is_tail( &mount_chain, node ) && !stop;
-    node = rtems_chain_next( node )
-  ) {
-    const rtems_filesystem_mount_table_entry_t *mt_entry =
-      (rtems_filesystem_mount_table_entry_t *) node;
+  int rv = 0;
 
-    stop = (*routine)( mt_entry, routine_arg );
+  if (
+    options == RTEMS_FILESYSTEM_READ_ONLY
+      || options == RTEMS_FILESYSTEM_READ_WRITE
+  ) {
+    rtems_filesystem_fsmount_me_t fsmount_me_h =
+      rtems_filesystem_get_mount_handler( filesystemtype );
+
+    if ( fsmount_me_h != NULL ) {
+      size_t target_length = 0;
+      rtems_filesystem_mount_table_entry_t *mt_entry = alloc_mount_table_entry(
+        source,
+        target,
+        filesystemtype,
+        &target_length
+      );
+
+      if ( mt_entry != NULL ) {
+        mt_entry->writeable = options == RTEMS_FILESYSTEM_READ_WRITE;
+
+        rv = (*fsmount_me_h)( mt_entry, data );
+        if ( rv == 0 ) {
+          if ( target != NULL ) {
+            rv = register_subordinate_file_system( mt_entry, target );
+          } else {
+            rv = register_root_file_system( mt_entry );
+          }
+
+          if ( rv != 0 ) {
+            (*mt_entry->mt_fs_root->location.ops->fsunmount_me_h)( mt_entry );
+          }
+        }
+
+        if ( rv != 0 ) {
+          free( mt_entry );
+        }
+      } else {
+        errno = ENOMEM;
+        rv = -1;
+      }
+    } else {
+      errno = EINVAL;
+      rv = -1;
+    }
+  } else {
+    errno = EINVAL;
+    rv = -1;
   }
-  rtems_libio_unlock();
 
-  return stop;
+  return rv;
 }