3  * mft.c - NTFS kernel mft record operations. Part of the Linux-NTFS project.
27  * map_mft_record_page - map the page in which a specific mft record resides
28  * @ni:		ntfs inode whose mft record page to map
30  * This maps the page in which the mft record of the ntfs inode @ni is situated
31  * and returns a pointer to the mft record within the mapped page.
48 	 * page of the wanted mft record. FIXME: We need to check for  in map_mft_record_page()
60 	/* If the wanted index is out of bounds the mft record doesn't exist. */  in map_mft_record_page()
65 			ntfs_error(vol->sb, "Attempt to read mft record 0x%lx, "  in map_mft_record_page()
82 		ntfs_error(vol->sb, "Mft record 0x%lx is corrupt.  "  in map_mft_record_page()
95  * map_mft_record - map, pin and lock an mft record
96  * @ni:		ntfs inode whose MFT record to map
101  * The page of the record is mapped using map_mft_record_page() before being
105  * record (it in turn calls read_cache_page() which reads it in from disk if
111  * and the post-read mst fixups on each mft record in the page have been
133  * The mft record is now ours and we return a pointer to it. You need to check
137  * NOTE: Caller is responsible for setting the mft record dirty before calling
139  * modified the mft record...
141  * A: No, the inode ones mean we want to change the mft record, not we want to
153 	/* Serialize access to this mft record. */  in map_mft_record()
167  * unmap_mft_record_page - unmap the page in which a specific mft record resides
168  * @ni:		ntfs inode whose mft record page to unmap
170  * This unmaps the page in which the mft record of the ntfs inode @ni is
192  * unmap_mft_record - release a mapped mft record
193  * @ni:		ntfs inode whose MFT record to unmap
196  * record and releases it for others to get hold of. We also release the ntfs
199  * NOTE: If caller has modified the mft record, it is imperative to set the mft
200  * record dirty BEFORE calling unmap_mft_record().
228  * Load the extent mft record @mref and attach it to its base inode @base_ni.
229  * Return the mapped extent mft record if IS_ERR(result) is false.  Otherwise
246 	ntfs_debug("Mapping extent mft record 0x%lx (base mft record 0x%lx).",  in map_extent_mft_record()
270 		/* We found the record; just have to map and return it. */  in map_extent_mft_record()
289 				"mft record, error code %ld.", -PTR_ERR(m));  in map_extent_mft_record()
292 	/* Record wasn't there. Get a new ntfs inode and initialize it. */  in map_extent_mft_record()
303 	/* Now map the record. */  in map_extent_mft_record()
362  * __mark_mft_record_dirty - set the mft record and the page containing it dirty
363  * @ni:		ntfs inode describing the mapped mft record
367  * Set the mapped (extent) mft record of the (base or extent) ntfs inode @ni,
368  * as well as the page containing the mft record, dirty.  Also, mark the base
369  * vfs inode dirty.  This ensures that any changes to the mft record are
406  * ntfs_sync_mft_mirror_umount - synchronise an mft record to the mft mirror
407  * @vol:	ntfs volume on which the mft record to synchronize resides
408  * @mft_no:	mft record number of mft record to synchronize
409  * @m:		mapped, mst protected (extent) mft record to synchronize
411  * Write the mapped, mst protected (extent) mft record @m with mft record
437  * ntfs_sync_mft_mirror - synchronize an mft record to the mft mirror
438  * @vol:	ntfs volume on which the mft record to synchronize resides
439  * @mft_no:	mft record number of mft record to synchronize
440  * @m:		mapped, mst protected (extent) mft record to synchronize
443  * Write the mapped, mst protected (extent) mft record @m with mft record
479 	/* Get the page containing the mirror copy of the mft record @m. */  in ntfs_sync_mft_mirror()
490 	/* Offset of the mft mirror record inside the page. */  in ntfs_sync_mft_mirror()
492 	/* The address in the page of the mirror copy of the mft record @m. */  in ntfs_sync_mft_mirror()
494 	/* Copy the mst protected mft record to the mirror. */  in ntfs_sync_mft_mirror()
518 		/* If the buffer is outside the mft record, skip it. */  in ntfs_sync_mft_mirror()
559 						"record 0x%lx because its "  in ntfs_sync_mft_mirror()
618 				"record 0x%lx!", mft_no);  in ntfs_sync_mft_mirror()
630  * write_mft_record_nolock - write out a mapped (extent) mft record
631  * @ni:		ntfs inode describing the mapped (extent) mft record
632  * @m:		mapped (extent) mft record to write
635  * Write the mapped (extent) mft record @m described by the (regular or extent)
636  * ntfs inode @ni to backing store.  If the mft record @m has a counterpart in
639  * We only write the mft record if the ntfs inode @ni is dirty and the first
640  * buffer belonging to its mft record is dirty, too.  We ignore the dirty state
644  * On success, clean the mft record and return 0.  On error, leave the mft
645  * record dirty and return -errno.
648  * However, if the mft record has a counterpart in the mft mirror and @sync is
649  * true, we write the mft record, wait for i/o completion, and only then write
651  * or the mft mirror will contain a self-consistent mft record @m.  If @sync is
654  * up with a self-consistent mft record in the case of a crash but if you asked
685 	 * for the mft record @m and the page it is in.  in write_mft_record_nolock()
698 		/* If the buffer is outside the mft record, skip it. */  in write_mft_record_nolock()
704 		 * If this block is not the first one in the record, we ignore  in write_mft_record_nolock()
709 			/* This block is the first one in the record. */  in write_mft_record_nolock()
746 				ntfs_error(vol->sb, "Cannot write mft record "  in write_mft_record_nolock()
811 		ntfs_error(vol->sb, "I/O error while writing mft record "  in write_mft_record_nolock()
832 		ntfs_error(vol->sb, "Not enough memory to write mft record.  "  in write_mft_record_nolock()
842  * ntfs_may_write_mft_record - check if an mft record may be written out
843  * @vol:	[IN]  ntfs volume on which the mft record to check resides
844  * @mft_no:	[IN]  mft record number of the mft record to check
845  * @m:		[IN]  mapped mft record to check
848  * Check if the mapped (base or extent) mft record @m with mft record number
850  * and possible the ntfs inode of the mft record is locked and the base vfs
855  * Return 'true' if the mft record may be written out and 'false' if not.
861  * record in read_cache_page() until we are finished with the write out.
865  * If the inode is found in icache we know the mft record must be a base mft
866  * record.  If it is dirty, we do not write it and return 'false' as the vfs
868  * cause the base mft record to be redirtied and written out again.  (We know
869  * the access time update will modify the base mft record because Windows
871  * mft record.)
873  * If the inode is in icache and not dirty, we attempt to lock the mft record
875  * record and we return 'false'.
877  * If we manage to obtain the lock we have exclusive access to the mft record,
878  * which also allows us safe writeout of the mft record.  We then set
881  * Note we cannot just lock the mft record and sleep while waiting for the lock
882  * because this would deadlock due to lock reversal (normally the mft record is
884  * when we try to lock the mft record).
888  * If the mft record is not a FILE record or it is a base mft record, we can
891  * We now know the mft record is an extent mft record.  We check if the inode
892  * corresponding to its base mft record is in icache and obtain a reference to
895  * We now have the base inode for the extent mft record.  We check if it has an
896  * ntfs inode for the extent mft record attached and if not it is safe to write
897  * the extent mft record and we return 'true'.
899  * The ntfs inode for the extent mft record is attached to the base inode so we
900  * attempt to lock the extent mft record and if we find the lock was already
901  * taken, it is not safe to write the extent mft record and we return 'false'.
904  * record, which also allows us safe writeout of the extent mft record.  We
905  * set the ntfs inode of the extent mft record clean and then set @locked_ni to
917  * appear if the mft record is reused for a new inode before it got written
937 	 * Check if the inode corresponding to this mft record is in the VFS  in ntfs_may_write_mft_record()
969 		/* If the inode is dirty, do not write this record. */  in ntfs_may_write_mft_record()
978 		/* The inode is not dirty, try to take the mft record lock. */  in ntfs_may_write_mft_record()
980 			ntfs_debug("Mft record 0x%lx is already locked, do "  in ntfs_may_write_mft_record()
986 		ntfs_debug("Managed to lock mft record 0x%lx, write it.",  in ntfs_may_write_mft_record()
989 		 * The write has to occur while we hold the mft record lock so  in ntfs_may_write_mft_record()
997 	/* Write the record if it is not a mft record (type "FILE"). */  in ntfs_may_write_mft_record()
999 		ntfs_debug("Mft record 0x%lx is not a FILE record, write it.",  in ntfs_may_write_mft_record()
1003 	/* Write the mft record if it is a base inode. */  in ntfs_may_write_mft_record()
1005 		ntfs_debug("Mft record 0x%lx is a base record, write it.",  in ntfs_may_write_mft_record()
1010 	 * This is an extent mft record.  Check if the inode corresponding to  in ntfs_may_write_mft_record()
1011 	 * its base mft record is in icache and obtain a reference to it if it  in ntfs_may_write_mft_record()
1015 	ntfs_debug("Mft record 0x%lx is an extent record.  Looking for base "  in ntfs_may_write_mft_record()
1027 		 * record.  in ntfs_may_write_mft_record()
1030 				"extent record.", na.mft_no);  in ntfs_may_write_mft_record()
1036 	 * corresponding to this extent mft record attached.  in ntfs_may_write_mft_record()
1043 		 * extent mft record.  in ntfs_may_write_mft_record()
1048 				"write the extent record.", na.mft_no);  in ntfs_may_write_mft_record()
1057 			 * mft record.  in ntfs_may_write_mft_record()
1065 	 * extent mft record.  in ntfs_may_write_mft_record()
1071 				"inode 0x%lx, write the extent record.",  in ntfs_may_write_mft_record()
1081 	 * Found the extent inode coresponding to this extent mft record.  in ntfs_may_write_mft_record()
1082 	 * Try to take the mft record lock.  in ntfs_may_write_mft_record()
1087 		ntfs_debug("Extent mft record 0x%lx is already locked, do "  in ntfs_may_write_mft_record()
1091 	ntfs_debug("Managed to lock extent mft record 0x%lx, write it.",  in ntfs_may_write_mft_record()
1097 	 * The write has to occur while we hold the mft record lock so return  in ntfs_may_write_mft_record()
1109  * @vol:	volume on which to search for a free mft record
1110  * @base_ni:	open base inode if allocating an extent mft record or NULL
1112  * Search for a free mft record in the mft bitmap attribute on the ntfs volume
1117  * If @base_ni is not NULL start the search at the mft record after the base
1118  * mft record @base_ni.
1120  * Return the free mft record on success and -errno on error.  An error code of
1137 	ntfs_debug("Searching for free mft record in the currently "  in ntfs_mft_bitmap_find_and_alloc_free_rec_nolock()
1172 	/* Loop until a free mft record is found. */  in ntfs_mft_bitmap_find_and_alloc_free_rec_nolock()
1217 							"allocated mft record "  in ntfs_mft_bitmap_find_and_alloc_free_rec_nolock()
1380 	 * Update the attribute record as well.  Note: @rl is the last  in ntfs_mft_bitmap_extend_allocation_nolock()
1385 		ntfs_error(vol->sb, "Failed to map mft record.");  in ntfs_mft_bitmap_extend_allocation_nolock()
1424 	/* Expand the attribute record if necessary. */  in ntfs_mft_bitmap_extend_allocation_nolock()
1431 					"record for mft bitmap attribute.");  in ntfs_mft_bitmap_extend_allocation_nolock()
1435 		// record or by starting a new extent in a new mft record or by  in ntfs_mft_bitmap_extend_allocation_nolock()
1436 		// moving other attributes out of this mft record.  in ntfs_mft_bitmap_extend_allocation_nolock()
1437 		// Note: It will need to be a special mft record and if none of  in ntfs_mft_bitmap_extend_allocation_nolock()
1439 		ntfs_error(vol->sb, "Not enough space in this mft record to "  in ntfs_mft_bitmap_extend_allocation_nolock()
1446 	/* Generate the mapping pairs array directly into the attr record. */  in ntfs_mft_bitmap_extend_allocation_nolock()
1459 	 * Reflect this in the ntfs_inode structure and the attribute record.  in ntfs_mft_bitmap_extend_allocation_nolock()
1544 					"record.%s", es);  in ntfs_mft_bitmap_extend_allocation_nolock()
1587 	/* Get the attribute record. */  in ntfs_mft_bitmap_extend_initialized_nolock()
1590 		ntfs_error(vol->sb, "Failed to map mft record.");  in ntfs_mft_bitmap_extend_initialized_nolock()
1642 		ntfs_error(vol->sb, "Failed to map mft record.%s", es);  in ntfs_mft_bitmap_extend_initialized_nolock()
1695  * worth of clusters or if not enough space for this by one mft record worth
1749 	/* Minimum allocation is one mft record worth of clusters. */  in ntfs_mft_data_extend_allocation_nolock()
1766 			ntfs_warning(vol->sb, "Cannot allocate mft record "  in ntfs_mft_data_extend_allocation_nolock()
1815 	/* Update the attribute record as well. */  in ntfs_mft_data_extend_allocation_nolock()
1818 		ntfs_error(vol->sb, "Failed to map mft record.");  in ntfs_mft_data_extend_allocation_nolock()
1856 	/* Expand the attribute record if necessary. */  in ntfs_mft_data_extend_allocation_nolock()
1863 					"record for mft data attribute.");  in ntfs_mft_data_extend_allocation_nolock()
1867 		// record or by starting a new extent in a new mft record or by  in ntfs_mft_data_extend_allocation_nolock()
1868 		// moving other attributes out of this mft record.  in ntfs_mft_data_extend_allocation_nolock()
1870 		// this extent is not required to find the mft record in  in ntfs_mft_data_extend_allocation_nolock()
1872 		// move an existing record away, insert ours in its place, and  in ntfs_mft_data_extend_allocation_nolock()
1873 		// then place the moved record into the newly allocated space  in ntfs_mft_data_extend_allocation_nolock()
1875 		// record appropriately.  This is rather complicated...  in ntfs_mft_data_extend_allocation_nolock()
1876 		ntfs_error(vol->sb, "Not enough space in this mft record to "  in ntfs_mft_data_extend_allocation_nolock()
1883 	/* Generate the mapping pairs array directly into the attr record. */  in ntfs_mft_data_extend_allocation_nolock()
1896 	 * Reflect this in the ntfs_inode structure and the attribute record.  in ntfs_mft_data_extend_allocation_nolock()
1976 					"record.%s", es);  in ntfs_mft_data_extend_allocation_nolock()
1995  * ntfs_mft_record_layout - layout an mft record into a memory buffer
1996  * @vol:	volume to which the mft record will belong
1997  * @mft_no:	mft reference specifying the mft record number
2000  * Layout an empty, unused mft record with the mft record number @mft_no into
2001  * the buffer @m.  The volume @vol is needed because the mft record structure
2003  * record will be used on.
2012 	ntfs_debug("Entering for mft record 0x%llx.", (long long)mft_no);  in ntfs_mft_record_layout()
2014 		ntfs_error(vol->sb, "Mft record number 0x%llx exceeds "  in ntfs_mft_record_layout()
2018 	/* Start by clearing the whole mft record to gives us a clean slate. */  in ntfs_mft_record_layout()
2038 		ntfs_warning(vol->sb, "Sector size is bigger than mft record "  in ntfs_mft_record_layout()
2076  * ntfs_mft_record_format - format an mft record on an ntfs volume
2077  * @vol:	volume on which to format the mft record
2078  * @mft_no:	mft record number to format
2080  * Format the mft record @mft_no in $MFT/$DATA, i.e. lay out an empty, unused
2081  * mft record into the appropriate place of the mft data attribute.  This is
2096 	ntfs_debug("Entering for mft record 0x%llx.", (long long)mft_no);  in ntfs_mft_record_format()
2099 	 * page of the wanted mft record.  in ntfs_mft_record_format()
2110 					"record 0x%llx.", (long long)mft_no);  in ntfs_mft_record_format()
2114 	/* Read, map, and pin the page containing the mft record. */  in ntfs_mft_record_format()
2117 		ntfs_error(vol->sb, "Failed to map page containing mft record "  in ntfs_mft_record_format()
2127 		ntfs_error(vol->sb, "Failed to layout mft record 0x%llx.",  in ntfs_mft_record_format()
2138 	 * Make sure the mft record is written out to disk.  We could use  in ntfs_mft_record_format()
2140 	 * unnecessary as ntfs_writepage() will write the dirty record anyway.  in ntfs_mft_record_format()
2149  * ntfs_mft_record_alloc - allocate an mft record on an ntfs volume
2150  * @vol:	[IN]  volume on which to allocate the mft record
2152  * @base_ni:	[IN]  open base inode if allocating an extent mft record or NULL
2153  * @mrec:	[OUT] on successful return this is the mapped mft record
2155  * Allocate an mft record in $MFT/$DATA of an open ntfs volume @vol.
2157  * If @base_ni is NULL make the mft record a base mft record, i.e. a file or
2163  * If @base_ni is not NULL make the allocated mft record an extent record,
2164  * allocate it starting at the mft record after the base mft record and attach
2170  * allocated mft record.  *@mrec is then set to the allocated, mapped, pinned,
2171  * and locked mft record.  If IS_ERR() is true, the function failed and the
2177  * To find a free mft record, we scan the mft bitmap for a zero bit.  To
2192  * record.  If there are no free records left in the range 24 to number of
2196  * is not sufficient space to do this, we try to extend by a single mft record
2197  * or one cluster, if cluster size is above the mft record size.
2200  * allocated mft record, incrementing mft data size and initialized size
2203  * mft records until we reach record 24 which we consider as the first free mft
2204  * record for use by normal files.
2215  *	- set the bit corresponding to the mft record being allocated in the
2217  *	- opened an ntfs_inode for the allocated mft record, and we will have
2219  *	  locked mft record.
2221  * On error, the volume will be left in a consistent state and no record will
2228  * the base mft record and an attribute list is used, it is very important that
2256 		ntfs_debug("Entering (allocating an extent mft record for "  in ntfs_mft_record_alloc()
2257 				"base mft record 0x%llx).",  in ntfs_mft_record_alloc()
2262 		ntfs_debug("Entering (allocating a base mft record).");  in ntfs_mft_record_alloc()
2276 		ntfs_debug("Found and allocated free record (#1), bit 0x%llx.",  in ntfs_mft_record_alloc()
2287 	 * so we can simply allocate the first unused mft record.  in ntfs_mft_record_alloc()
2304 		ntfs_debug("Found free record (#2), bit 0x%llx.",  in ntfs_mft_record_alloc()
2310 	 * mft record that we can allocate.  in ntfs_mft_record_alloc()
2311 	 * Note: The smallest mft record we allocate is mft record 24.  in ntfs_mft_record_alloc()
2363 	ntfs_debug("Found free record (#3), bit 0x%llx.", (long long)bit);  in ntfs_mft_record_alloc()
2365 	/* @bit is the found free mft record, allocate it in the mft bitmap. */  in ntfs_mft_record_alloc()
2379 	 * will impact decisions for mft bitmap and mft record allocation done  in ntfs_mft_record_alloc()
2381 	 * parallel allocation could allocate the same mft record as this one.  in ntfs_mft_record_alloc()
2388 		ntfs_debug("Allocated mft record already initialized.");  in ntfs_mft_record_alloc()
2391 	ntfs_debug("Initializing allocated mft record.");  in ntfs_mft_record_alloc()
2393 	 * The mft record is outside the initialized data.  Extend the mft data  in ntfs_mft_record_alloc()
2394 	 * attribute until it covers the allocated record.  The loop is only  in ntfs_mft_record_alloc()
2424 	 * the allocated mft record, formatting the mft records allong the way.  in ntfs_mft_record_alloc()
2427 	 * record itself in one fell swoop later on.  in ntfs_mft_record_alloc()
2441 		ntfs_debug("Initializing mft record 0x%llx.",  in ntfs_mft_record_alloc()
2445 			ntfs_error(vol->sb, "Failed to format mft record.");  in ntfs_mft_record_alloc()
2453 	/* Update the mft data attribute record to reflect the new sizes. */  in ntfs_mft_record_alloc()
2456 		ntfs_error(vol->sb, "Failed to map mft record.");  in ntfs_mft_record_alloc()
2489 	ntfs_debug("Status of mft data after mft record initialization: "  in ntfs_mft_record_alloc()
2502 	 * allocated mft record still needs to be marked as in use to match the  in ntfs_mft_record_alloc()
2504 	 * this mft record is not referenced from anywhere yet and the fact  in ntfs_mft_record_alloc()
2510 	 * We now have allocated and initialized the mft record.  Calculate the  in ntfs_mft_record_alloc()
2511 	 * index of and the offset within the page cache page the record is in.  in ntfs_mft_record_alloc()
2515 	/* Read, map, and pin the page containing the mft record. */  in ntfs_mft_record_alloc()
2519 				"mft record 0x%llx.", (long long)bit);  in ntfs_mft_record_alloc()
2527 	/* If we just formatted the mft record no need to do it again. */  in ntfs_mft_record_alloc()
2529 		/* Sanity check that the mft record is really not in use. */  in ntfs_mft_record_alloc()
2532 			ntfs_error(vol->sb, "Mft record 0x%llx was marked "  in ntfs_mft_record_alloc()
2545 		 * We need to (re-)format the mft record, preserving the  in ntfs_mft_record_alloc()
2549 		 * wrong with the previous mft record.  in ntfs_mft_record_alloc()
2556 					"record 0x%llx.", (long long)bit);  in ntfs_mft_record_alloc()
2567 	/* Set the mft record itself in use. */  in ntfs_mft_record_alloc()
2577 		 * Setup the base mft record in the extent mft record.  This  in ntfs_mft_record_alloc()
2578 		 * completes initialization of the allocated extent mft record  in ntfs_mft_record_alloc()
2584 		 * Allocate an extent inode structure for the new mft record,  in ntfs_mft_record_alloc()
2586 		 * its, i.e. the allocated, mft record.  in ntfs_mft_record_alloc()
2591 					"mft record 0x%llx.", (long long)bit);  in ntfs_mft_record_alloc()
2593 			/* Set the mft record itself not in use. */  in ntfs_mft_record_alloc()
2597 			/* Make sure the mft record is written out to disk. */  in ntfs_mft_record_alloc()
2605 		 * Make sure the allocated mft record is written out to disk.  in ntfs_mft_record_alloc()
2608 		 * record (e.g. at a minimum a new attribute will be added to  in ntfs_mft_record_alloc()
2609 		 * the mft record.  in ntfs_mft_record_alloc()
2621 		 * is set to 1 but the mft record->link_count is 0.  The caller  in ntfs_mft_record_alloc()
2627 			/* Set the mft record itself not in use. */  in ntfs_mft_record_alloc()
2631 			/* Make sure the mft record is written out to disk. */  in ntfs_mft_record_alloc()
2695 		 * Manually map, pin, and lock the mft record as we already  in ntfs_mft_record_alloc()
2703 		 * Make sure the allocated mft record is written out to disk.  in ntfs_mft_record_alloc()
2708 		 * that anyway after finishing with the new mft record (e.g. at  in ntfs_mft_record_alloc()
2710 		 * record.  in ntfs_mft_record_alloc()
2722 	 * Return the opened, allocated inode of the allocated mft record as  in ntfs_mft_record_alloc()
2723 	 * well as the mapped, pinned, and locked mft record.  in ntfs_mft_record_alloc()
2746 	ntfs_warning(vol->sb, "Cannot allocate mft record because the maximum "  in ntfs_mft_record_alloc()
2753  * ntfs_extent_mft_record_free - free an extent mft record on an ntfs volume
2754  * @ni:		ntfs inode of the mapped extent mft record to free
2755  * @m:		mapped extent mft record of the ntfs inode @ni
2757  * Free the mapped extent mft record @m of the extent ntfs inode @ni.
2759  * Note that this function unmaps the mft record and closes and destroys @ni
2769  * caller no longer wants to reference this mft record so it does not matter to
2833 	/* Mark the mft record as not in use. */  in ntfs_extent_mft_record_free()
2848 	 * record to be freed is guaranteed to do it already.  in ntfs_extent_mft_record_free()
2853 		ntfs_error(vol->sb, "Failed to write mft record 0x%lx, not "  in ntfs_extent_mft_record_free()
2862 	/* Clear the bit in the $MFT/$BITMAP corresponding to this record. */  in ntfs_extent_mft_record_free()