NTFS: Fix a bug in fs/ntfs/inode.c::ntfs_read_locked_index_inode() where we

forgot to update a temporary variable so loading index inodes which
      have an index allocation attribute failed.

Signed-off-by: Anton Altaparmakov <aia21@cantab.net>
This commit is contained in:
Anton Altaparmakov
2006-03-23 16:18:23 +00:00
parent 2c2c8c1c21
commit a778f21732
3 changed files with 15 additions and 19 deletions

View File

@@ -651,10 +651,7 @@ err_out:
* fs/ntfs/aops.c::mark_ntfs_record_dirty().
*
* On success, clean the mft record and return 0. On error, leave the mft
* record dirty and return -errno. The caller should call make_bad_inode() on
* the base inode to ensure no more access happens to this inode. We do not do
* it here as the caller may want to finish writing other extent mft records
* first to minimize on-disk metadata inconsistencies.
* record dirty and return -errno.
*
* NOTE: We always perform synchronous i/o and ignore the @sync parameter.
* However, if the mft record has a counterpart in the mft mirror and @sync is