xfs_dir2_leaf.c

来自「Linux Kernel 2.6.9 for OMAP1710」· C语言 代码 · 共 1,897 行 · 第 1/4 页

	 * It's probably not at the right location, so we'll have to
	 * shift some up or down first.
	 */
	else {
		/*
		 * If we didn't compact before, we need to find the nearest
		 * stale entries before and after our insertion point.
		 */
		if (compact == 0) {
			/*
			 * Find the first stale entry before the insertion
			 * point, if any.
			 */
			for (lowstale = index - 1;
			     lowstale >= 0 &&
				INT_GET(leaf->ents[lowstale].address, ARCH_CONVERT) !=
				XFS_DIR2_NULL_DATAPTR;
			     lowstale--)
				continue;
			/*
			 * Find the next stale entry at or after the insertion
			 * point, if any.   Stop if we go so far that the
			 * lowstale entry would be better.
			 */
			for (highstale = index;
			     highstale < INT_GET(leaf->hdr.count, ARCH_CONVERT) &&
				INT_GET(leaf->ents[highstale].address, ARCH_CONVERT) !=
				XFS_DIR2_NULL_DATAPTR &&
				(lowstale < 0 ||
				 index - lowstale - 1 >= highstale - index);
			     highstale++)
				continue;
		}
		/*
		 * If the low one is better, use it.
		 */
		if (lowstale >= 0 &&
		    (highstale == INT_GET(leaf->hdr.count, ARCH_CONVERT) ||
		     index - lowstale - 1 < highstale - index)) {
			ASSERT(index - lowstale - 1 >= 0);
			ASSERT(INT_GET(leaf->ents[lowstale].address, ARCH_CONVERT) ==
			       XFS_DIR2_NULL_DATAPTR);
			/*
			 * Copy entries up to cover the stale entry
			 * and make room for the new entry.
			 */
			if (index - lowstale - 1 > 0)
				memmove(&leaf->ents[lowstale],
					&leaf->ents[lowstale + 1],
					(index - lowstale - 1) * sizeof(*lep));
			lep = &leaf->ents[index - 1];
			lfloglow = MIN(lowstale, lfloglow);
			lfloghigh = MAX(index - 1, lfloghigh);
		}
		/*
		 * The high one is better, so use that one.
		 */
		else {
			ASSERT(highstale - index >= 0);
			ASSERT(INT_GET(leaf->ents[highstale].address, ARCH_CONVERT) ==
			       XFS_DIR2_NULL_DATAPTR);
			/*
			 * Copy entries down to copver the stale entry
			 * and make room for the new entry.
			 */
			if (highstale - index > 0)
				memmove(&leaf->ents[index + 1],
					&leaf->ents[index],
					(highstale - index) * sizeof(*lep));
			lep = &leaf->ents[index];
			lfloglow = MIN(index, lfloglow);
			lfloghigh = MAX(highstale, lfloghigh);
		}
		INT_MOD(leaf->hdr.stale, ARCH_CONVERT, -1);
	}
	/*
	 * Fill in the new leaf entry.
	 */
	INT_SET(lep->hashval, ARCH_CONVERT, args->hashval);
	INT_SET(lep->address, ARCH_CONVERT, XFS_DIR2_DB_OFF_TO_DATAPTR(mp, use_block, INT_GET(*tagp, ARCH_CONVERT)));
	/*
	 * Log the leaf fields and give up the buffers.
	 */
	xfs_dir2_leaf_log_header(tp, lbp);
	xfs_dir2_leaf_log_ents(tp, lbp, lfloglow, lfloghigh);
	xfs_dir2_leaf_check(dp, lbp);
	xfs_da_buf_done(lbp);
	xfs_dir2_data_check(dp, dbp);
	xfs_da_buf_done(dbp);
	return 0;
}

#ifdef DEBUG
/*
 * Check the internal consistency of a leaf1 block.
 * Pop an assert if something is wrong.
 */
void
xfs_dir2_leaf_check(
	xfs_inode_t		*dp,		/* incore directory inode */
	xfs_dabuf_t		*bp)		/* leaf's buffer */
{
	int			i;		/* leaf index */
	xfs_dir2_leaf_t		*leaf;		/* leaf structure */
	xfs_dir2_leaf_tail_t	*ltp;		/* leaf tail pointer */
	xfs_mount_t		*mp;		/* filesystem mount point */
	int			stale;		/* count of stale leaves */

	leaf = bp->data;
	mp = dp->i_mount;
	ASSERT(INT_GET(leaf->hdr.info.magic, ARCH_CONVERT) == XFS_DIR2_LEAF1_MAGIC);
	/*
	 * This value is not restrictive enough.
	 * Should factor in the size of the bests table as well.
	 * We can deduce a value for that from di_size.
	 */
	ASSERT(INT_GET(leaf->hdr.count, ARCH_CONVERT) <= XFS_DIR2_MAX_LEAF_ENTS(mp));
	ltp = XFS_DIR2_LEAF_TAIL_P(mp, leaf);
	/*
	 * Leaves and bests don't overlap.
	 */
	ASSERT((char *)&leaf->ents[INT_GET(leaf->hdr.count, ARCH_CONVERT)] <=
	       (char *)XFS_DIR2_LEAF_BESTS_P_ARCH(ltp, ARCH_CONVERT));
	/*
	 * Check hash value order, count stale entries.
	 */
	for (i = stale = 0; i < INT_GET(leaf->hdr.count, ARCH_CONVERT); i++) {
		if (i + 1 < INT_GET(leaf->hdr.count, ARCH_CONVERT))
			ASSERT(INT_GET(leaf->ents[i].hashval, ARCH_CONVERT) <=
			       INT_GET(leaf->ents[i + 1].hashval, ARCH_CONVERT));
		if (INT_GET(leaf->ents[i].address, ARCH_CONVERT) == XFS_DIR2_NULL_DATAPTR)
			stale++;
	}
	ASSERT(INT_GET(leaf->hdr.stale, ARCH_CONVERT) == stale);
}
#endif	/* DEBUG */

/*
 * Compact out any stale entries in the leaf.
 * Log the header and changed leaf entries, if any.
 */
void
xfs_dir2_leaf_compact(
	xfs_da_args_t	*args,		/* operation arguments */
	xfs_dabuf_t	*bp)		/* leaf buffer */
{
	int		from;		/* source leaf index */
	xfs_dir2_leaf_t	*leaf;		/* leaf structure */
	int		loglow;		/* first leaf entry to log */
	int		to;		/* target leaf index */

	leaf = bp->data;
	if (INT_ISZERO(leaf->hdr.stale, ARCH_CONVERT)) {
		return;
	}
	/*
	 * Compress out the stale entries in place.
	 */
	for (from = to = 0, loglow = -1; from < INT_GET(leaf->hdr.count, ARCH_CONVERT); from++) {
		if (INT_GET(leaf->ents[from].address, ARCH_CONVERT) == XFS_DIR2_NULL_DATAPTR)
			continue;
		/*
		 * Only actually copy the entries that are different.
		 */
		if (from > to) {
			if (loglow == -1)
				loglow = to;
			leaf->ents[to] = leaf->ents[from];
		}
		to++;
	}
	/*
	 * Update and log the header, log the leaf entries.
	 */
	ASSERT(INT_GET(leaf->hdr.stale, ARCH_CONVERT) == from - to);
	INT_MOD(leaf->hdr.count, ARCH_CONVERT, -(INT_GET(leaf->hdr.stale, ARCH_CONVERT)));
	INT_ZERO(leaf->hdr.stale, ARCH_CONVERT);
	xfs_dir2_leaf_log_header(args->trans, bp);
	if (loglow != -1)
		xfs_dir2_leaf_log_ents(args->trans, bp, loglow, to - 1);
}

/*
 * Compact the leaf entries, removing stale ones.
 * Leave one stale entry behind - the one closest to our
 * insertion index - and the caller will shift that one to our insertion
 * point later.
 * Return new insertion index, where the remaining stale entry is,
 * and leaf logging indices.
 */
void
xfs_dir2_leaf_compact_x1(
	xfs_dabuf_t	*bp,		/* leaf buffer */
	int		*indexp,	/* insertion index */
	int		*lowstalep,	/* out: stale entry before us */
	int		*highstalep,	/* out: stale entry after us */
	int		*lowlogp,	/* out: low log index */
	int		*highlogp)	/* out: high log index */
{
	int		from;		/* source copy index */
	int		highstale;	/* stale entry at/after index */
	int		index;		/* insertion index */
	int		keepstale;	/* source index of kept stale */
	xfs_dir2_leaf_t	*leaf;		/* leaf structure */
	int		lowstale;	/* stale entry before index */
	int		newindex=0;	/* new insertion index */
	int		to;		/* destination copy index */

	leaf = bp->data;
	ASSERT(INT_GET(leaf->hdr.stale, ARCH_CONVERT) > 1);
	index = *indexp;
	/*
	 * Find the first stale entry before our index, if any.
	 */
	for (lowstale = index - 1;
	     lowstale >= 0 &&
		INT_GET(leaf->ents[lowstale].address, ARCH_CONVERT) != XFS_DIR2_NULL_DATAPTR;
	     lowstale--)
		continue;
	/*
	 * Find the first stale entry at or after our index, if any.
	 * Stop if the answer would be worse than lowstale.
	 */
	for (highstale = index;
	     highstale < INT_GET(leaf->hdr.count, ARCH_CONVERT) &&
		INT_GET(leaf->ents[highstale].address, ARCH_CONVERT) != XFS_DIR2_NULL_DATAPTR &&
		(lowstale < 0 || index - lowstale > highstale - index);
	     highstale++)
		continue;
	/*
	 * Pick the better of lowstale and highstale.
	 */
	if (lowstale >= 0 &&
	    (highstale == INT_GET(leaf->hdr.count, ARCH_CONVERT) ||
	     index - lowstale <= highstale - index))
		keepstale = lowstale;
	else
		keepstale = highstale;
	/*
	 * Copy the entries in place, removing all the stale entries
	 * except keepstale.
	 */
	for (from = to = 0; from < INT_GET(leaf->hdr.count, ARCH_CONVERT); from++) {
		/*
		 * Notice the new value of index.
		 */
		if (index == from)
			newindex = to;
		if (from != keepstale &&
		    INT_GET(leaf->ents[from].address, ARCH_CONVERT) == XFS_DIR2_NULL_DATAPTR) {
			if (from == to)
				*lowlogp = to;
			continue;
		}
		/*
		 * Record the new keepstale value for the insertion.
		 */
		if (from == keepstale)
			lowstale = highstale = to;
		/*
		 * Copy only the entries that have moved.
		 */
		if (from > to)
			leaf->ents[to] = leaf->ents[from];
		to++;
	}
	ASSERT(from > to);
	/*
	 * If the insertion point was past the last entry,
	 * set the new insertion point accordingly.
	 */
	if (index == from)
		newindex = to;
	*indexp = newindex;
	/*
	 * Adjust the leaf header values.
	 */
	INT_MOD(leaf->hdr.count, ARCH_CONVERT, -(from - to));
	INT_SET(leaf->hdr.stale, ARCH_CONVERT, 1);
	/*
	 * Remember the low/high stale value only in the "right"
	 * direction.
	 */
	if (lowstale >= newindex)
		lowstale = -1;
	else
		highstale = INT_GET(leaf->hdr.count, ARCH_CONVERT);
	*highlogp = INT_GET(leaf->hdr.count, ARCH_CONVERT) - 1;
	*lowstalep = lowstale;
	*highstalep = highstale;
}

/*
 * Getdents (readdir) for leaf and node directories.
 * This reads the data blocks only, so is the same for both forms.
 */
int						/* error */
xfs_dir2_leaf_getdents(
	xfs_trans_t		*tp,		/* transaction pointer */
	xfs_inode_t		*dp,		/* incore directory inode */
	uio_t			*uio,		/* I/O control & vectors */
	int			*eofp,		/* out: reached end of dir */
	xfs_dirent_t		*dbp,		/* caller's buffer */
	xfs_dir2_put_t		put)		/* ABI formatting routine */
{
	xfs_dabuf_t		*bp;		/* data block buffer */
	int			byteoff;	/* offset in current block */
	xfs_dir2_db_t		curdb;		/* db for current block */
	xfs_dir2_off_t		curoff;		/* current overall offset */
	xfs_dir2_data_t		*data;		/* data block structure */
	xfs_dir2_data_entry_t	*dep;		/* data entry */
	xfs_dir2_data_unused_t	*dup;		/* unused entry */
	int			eof;		/* reached end of directory */
	int			error=0;		/* error return value */
	int			i;		/* temporary loop index */
	int			j;		/* temporary loop index */
	int			length;		/* temporary length value */
	xfs_bmbt_irec_t		*map;		/* map vector for blocks */
	xfs_extlen_t		map_blocks;	/* number of fsbs in map */
	xfs_dablk_t		map_off;	/* last mapped file offset */
	int			map_size;	/* total entries in *map */
	int			map_valid;	/* valid entries in *map */
	xfs_mount_t		*mp;		/* filesystem mount point */
	xfs_dir2_off_t		newoff;		/* new curoff after new blk */
	int			nmap;		/* mappings to ask xfs_bmapi */
	xfs_dir2_put_args_t	p;		/* formatting arg bundle */
	char			*ptr=NULL;		/* pointer to current data */
	int			ra_current;	/* number of read-ahead blks */
	int			ra_index;	/* *map index for read-ahead */
	int			ra_offset;	/* map entry offset for ra */
	int			ra_want;	/* readahead count wanted */

	/*
	 * If the offset is at or past the largest allowed value,
	 * give up right away, return eof.
	 */
	if (uio->uio_offset >= XFS_DIR2_MAX_DATAPTR) {
		*eofp = 1;
		return 0;
	}
	mp = dp->i_mount;
	/*
	 * Setup formatting arguments.
	 */
	p.dbp = dbp;
	p.put = put;
	p.uio = uio;
	/*
	 * Set up to bmap a number of blocks based on the caller's
	 * buffer size, the directory block size, and the filesystem
	 * block size.
	 */
	map_size =
		howmany(uio->uio_resid + mp->m_dirblksize,
			mp->m_sb.sb_blocksize);
	map = kmem_alloc(map_size * sizeof(*map), KM_SLEEP);
	map_valid = ra_index = ra_offset = ra_current = map_blocks = 0;
	bp = NULL;
	eof = 1;
	/*
	 * Inside the loop we keep the main offset value as a byte offset
	 * in the directory file.
	 */
	curoff = XFS_DIR2_DATAPTR_TO_BYTE(mp, uio->uio_offset);
	/*
	 * Force this conversion through db so we truncate the offset
	 * down to get the start of the data block.
	 */
	map_off = XFS_DIR2_DB_TO_DA(mp, XFS_DIR2_BYTE_TO_DB(mp, curoff));
	/*
	 * Loop over directory entries until we reach the end offset.
	 * Get more blocks and readahead as necessary.
	 */
	while (curoff < XFS_DIR2_LEAF_OFFSET) {
		/*
		 * If we have no buffer, or we're off the end of the
		 * current buffer, need to get another one.
		 */
		if (!bp || ptr >= (char *)bp->data + mp->m_dirblksize) {
			/*
			 * If we have a buffer, we need to release it and
			 * take it out of the mapping.
			 */
			if (bp) {
				xfs_da_brelse(tp, bp);
				bp = NULL;
				map_blocks -= mp->m_dirblkfsbs;
				/*
				 * Loop to get rid of the extents for the
				 * directory block.
				 */
				for (i = mp->m_dirblkfsbs; i > 0; ) {
					j = MIN((int)map->br_blockcount, i);
					map->br_blockcount -= j;
					map->br_startblock += j;
					map->br_startoff += j;
					/*
					 * If mapping is done, pitch it from
					 * the table.
					 */
					if (!map->br_blockcount && --map_valid)
						memmove(&map[0], &map[1],
							sizeof(map[0]) *
							map_valid);
					i -= j;
				}
			}
			/*
			 * Recalculate the readahead blocks wanted.
			 */
			ra_want = howmany(uio->uio_resid + mp->m_dirblksize,
					  mp->m_sb.sb_blocksize) - 1;
			/*
			 * If we don't have as many as we want, and we haven't
			 * run out of data blocks, get some more mappings.
			 */
			if (1 + ra_want > map_blocks &&
			    map_off <
			    XFS_DIR2_BYTE_TO_DA(mp, XFS_DIR2_LEAF_OFFSET)) {
				/*
				 * Get more bmaps, fill in after the ones
				 * we already have in the table.
				 */
				nmap = map_size - map_valid;
				error = xfs_bmapi(tp, dp,
					map_off,
					XFS_DIR2_BYTE_TO_DA(mp,
						XFS_DIR2_LEAF_OFFSET) - map_off,
					XFS_BMAPI_METADATA, NULL, 0,
					&map[map_valid], &nmap, NULL);
				/*
				 * Don't know if we should ignore this or
				 * try to return an error.
				 * The trouble with returning errors
				 * is that readdir will just stop without
				 * actually passing the error through.
				 */
				if (error)
					break;	/* XXX */
				/*
				 * If we got all the mappings we asked for,
				 * set the final map offset based on the
				 * last bmap value received.
				 * Otherwise, we've reached the end.
				 */
				if (nmap == map_size - map_valid)
					map_off =
					map[map_valid + nmap - 1].br_startoff +
					map[map_valid + nmap - 1].br_blockcount;
				else
					map_off =
						XFS_DIR2_BYTE_TO_DA(mp,
							XFS_DIR2_LEAF_OFFSET);
				/*
				 * Look for holes in the mapping, and
				 * eliminate them.  Count up the valid blocks.
				 */
				for (i = map_valid; i < map_valid + nmap; ) {
					if (map[i].br_startblock ==
					    HOLESTARTBLOCK) {
						nmap--;
						length = map_valid + nmap - i;
						if (length)
							memmove(&map[i],
								&map[i + 1],
								sizeof(map[i]) *
								length);
					} else {
						map_blocks +=
							map[i].br_blockcount;
						i++;
					}
				}
				map_valid += nmap;
			}