xfs_rtalloc.c

linux 内核源代码

	xfs_sb_t	*sbp;		/* old superblock */
	xfs_fsblock_t	sumbno;		/* summary block number */
	xfs_trans_t	*tp;		/* transaction pointer */

	sbp = &mp->m_sb;
	/*
	 * Initial error checking.
	 */
	if (mp->m_rtdev_targp == NULL || mp->m_rbmip == NULL ||
	    (nrblocks = in->newblocks) <= sbp->sb_rblocks ||
	    (sbp->sb_rblocks && (in->extsize != sbp->sb_rextsize)))
		return XFS_ERROR(EINVAL);
	if ((error = xfs_sb_validate_fsb_count(sbp, nrblocks)))
		return error;
	/*
	 * Read in the last block of the device, make sure it exists.
	 */
	error = xfs_read_buf(mp, mp->m_rtdev_targp,
			XFS_FSB_TO_BB(mp, nrblocks - 1),
			XFS_FSB_TO_BB(mp, 1), 0, &bp);
	if (error)
		return error;
	ASSERT(bp);
	xfs_buf_relse(bp);
	/*
	 * Calculate new parameters.  These are the final values to be reached.
	 */
	nrextents = nrblocks;
	do_div(nrextents, in->extsize);
	nrbmblocks = howmany_64(nrextents, NBBY * sbp->sb_blocksize);
	nrextslog = xfs_highbit32(nrextents);
	nrsumlevels = nrextslog + 1;
	nrsumsize = (uint)sizeof(xfs_suminfo_t) * nrsumlevels * nrbmblocks;
	nrsumblocks = XFS_B_TO_FSB(mp, nrsumsize);
	nrsumsize = XFS_FSB_TO_B(mp, nrsumblocks);
	/*
	 * New summary size can't be more than half the size of
	 * the log.  This prevents us from getting a log overflow,
	 * since we'll log basically the whole summary file at once.
	 */
	if (nrsumblocks > (mp->m_sb.sb_logblocks >> 1))
		return XFS_ERROR(EINVAL);
	/*
	 * Get the old block counts for bitmap and summary inodes.
	 * These can't change since other growfs callers are locked out.
	 */
	rbmblocks = XFS_B_TO_FSB(mp, mp->m_rbmip->i_d.di_size);
	rsumblocks = XFS_B_TO_FSB(mp, mp->m_rsumip->i_d.di_size);
	/*
	 * Allocate space to the bitmap and summary files, as necessary.
	 */
	if ((error = xfs_growfs_rt_alloc(mp, rbmblocks, nrbmblocks,
			mp->m_sb.sb_rbmino)))
		return error;
	if ((error = xfs_growfs_rt_alloc(mp, rsumblocks, nrsumblocks,
			mp->m_sb.sb_rsumino)))
		return error;
	/*
	 * Allocate a new (fake) mount/sb.
	 */
	nmp = kmem_alloc(sizeof(*nmp), KM_SLEEP);
	/*
	 * Loop over the bitmap blocks.
	 * We will do everything one bitmap block at a time.
	 * Skip the current block if it is exactly full.
	 * This also deals with the case where there were no rtextents before.
	 */
	for (bmbno = sbp->sb_rbmblocks -
		     ((sbp->sb_rextents & ((1 << mp->m_blkbit_log) - 1)) != 0);
	     bmbno < nrbmblocks;
	     bmbno++) {
		*nmp = *mp;
		nsbp = &nmp->m_sb;
		/*
		 * Calculate new sb and mount fields for this round.
		 */
		nsbp->sb_rextsize = in->extsize;
		nsbp->sb_rbmblocks = bmbno + 1;
		nsbp->sb_rblocks =
			XFS_RTMIN(nrblocks,
				  nsbp->sb_rbmblocks * NBBY *
				  nsbp->sb_blocksize * nsbp->sb_rextsize);
		nsbp->sb_rextents = nsbp->sb_rblocks;
		do_div(nsbp->sb_rextents, nsbp->sb_rextsize);
		nsbp->sb_rextslog = xfs_highbit32(nsbp->sb_rextents);
		nrsumlevels = nmp->m_rsumlevels = nsbp->sb_rextslog + 1;
		nrsumsize =
			(uint)sizeof(xfs_suminfo_t) * nrsumlevels *
			nsbp->sb_rbmblocks;
		nrsumblocks = XFS_B_TO_FSB(mp, nrsumsize);
		nmp->m_rsumsize = nrsumsize = XFS_FSB_TO_B(mp, nrsumblocks);
		/*
		 * Start a transaction, get the log reservation.
		 */
		tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_FREE);
		cancelflags = 0;
		if ((error = xfs_trans_reserve(tp, 0,
				XFS_GROWRTFREE_LOG_RES(nmp), 0, 0, 0)))
			break;
		/*
		 * Lock out other callers by grabbing the bitmap inode lock.
		 */
		if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0,
						XFS_ILOCK_EXCL, &ip)))
			break;
		ASSERT(ip == mp->m_rbmip);
		/*
		 * Update the bitmap inode's size.
		 */
		mp->m_rbmip->i_d.di_size =
			nsbp->sb_rbmblocks * nsbp->sb_blocksize;
		xfs_trans_log_inode(tp, mp->m_rbmip, XFS_ILOG_CORE);
		cancelflags |= XFS_TRANS_ABORT;
		/*
		 * Get the summary inode into the transaction.
		 */
		if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rsumino, 0,
						XFS_ILOCK_EXCL, &ip)))
			break;
		ASSERT(ip == mp->m_rsumip);
		/*
		 * Update the summary inode's size.
		 */
		mp->m_rsumip->i_d.di_size = nmp->m_rsumsize;
		xfs_trans_log_inode(tp, mp->m_rsumip, XFS_ILOG_CORE);
		/*
		 * Copy summary data from old to new sizes.
		 * Do this when the real size (not block-aligned) changes.
		 */
		if (sbp->sb_rbmblocks != nsbp->sb_rbmblocks ||
		    mp->m_rsumlevels != nmp->m_rsumlevels) {
			error = xfs_rtcopy_summary(mp, nmp, tp);
			if (error)
				break;
		}
		/*
		 * Update superblock fields.
		 */
		if (nsbp->sb_rextsize != sbp->sb_rextsize)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_REXTSIZE,
				nsbp->sb_rextsize - sbp->sb_rextsize);
		if (nsbp->sb_rbmblocks != sbp->sb_rbmblocks)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_RBMBLOCKS,
				nsbp->sb_rbmblocks - sbp->sb_rbmblocks);
		if (nsbp->sb_rblocks != sbp->sb_rblocks)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_RBLOCKS,
				nsbp->sb_rblocks - sbp->sb_rblocks);
		if (nsbp->sb_rextents != sbp->sb_rextents)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_REXTENTS,
				nsbp->sb_rextents - sbp->sb_rextents);
		if (nsbp->sb_rextslog != sbp->sb_rextslog)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_REXTSLOG,
				nsbp->sb_rextslog - sbp->sb_rextslog);
		/*
		 * Free new extent.
		 */
		bp = NULL;
		error = xfs_rtfree_range(nmp, tp, sbp->sb_rextents,
			nsbp->sb_rextents - sbp->sb_rextents, &bp, &sumbno);
		if (error)
			break;
		/*
		 * Mark more blocks free in the superblock.
		 */
		xfs_trans_mod_sb(tp, XFS_TRANS_SB_FREXTENTS,
			nsbp->sb_rextents - sbp->sb_rextents);
		/*
		 * Update mp values into the real mp structure.
		 */
		mp->m_rsumlevels = nrsumlevels;
		mp->m_rsumsize = nrsumsize;
		/*
		 * Commit the transaction.
		 */
		xfs_trans_commit(tp, 0);
	}

	if (error)
		xfs_trans_cancel(tp, cancelflags);

	/*
	 * Free the fake mp structure.
	 */
	kmem_free(nmp, sizeof(*nmp));

	return error;
}

/*
 * Allocate an extent in the realtime subvolume, with the usual allocation
 * parameters.  The length units are all in realtime extents, as is the
 * result block number.
 */
int					/* error */
xfs_rtallocate_extent(
	xfs_trans_t	*tp,		/* transaction pointer */
	xfs_rtblock_t	bno,		/* starting block number to allocate */
	xfs_extlen_t	minlen,		/* minimum length to allocate */
	xfs_extlen_t	maxlen,		/* maximum length to allocate */
	xfs_extlen_t	*len,		/* out: actual length allocated */
	xfs_alloctype_t	type,		/* allocation type XFS_ALLOCTYPE... */
	int		wasdel,		/* was a delayed allocation extent */
	xfs_extlen_t	prod,		/* extent product factor */
	xfs_rtblock_t	*rtblock)	/* out: start block allocated */
{
	int		error;		/* error value */
	xfs_inode_t	*ip;		/* inode for bitmap file */
	xfs_mount_t	*mp;		/* file system mount structure */
	xfs_rtblock_t	r;		/* result allocated block */
	xfs_fsblock_t	sb;		/* summary file block number */
	xfs_buf_t	*sumbp;		/* summary file block buffer */

	ASSERT(minlen > 0 && minlen <= maxlen);
	mp = tp->t_mountp;
	/*
	 * If prod is set then figure out what to do to minlen and maxlen.
	 */
	if (prod > 1) {
		xfs_extlen_t	i;

		if ((i = maxlen % prod))
			maxlen -= i;
		if ((i = minlen % prod))
			minlen += prod - i;
		if (maxlen < minlen) {
			*rtblock = NULLRTBLOCK;
			return 0;
		}
	}
	/*
	 * Lock out other callers by grabbing the bitmap inode lock.
	 */
	if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0,
					XFS_ILOCK_EXCL, &ip)))
		return error;
	sumbp = NULL;
	/*
	 * Allocate by size, or near another block, or exactly at some block.
	 */
	switch (type) {
	case XFS_ALLOCTYPE_ANY_AG:
		error = xfs_rtallocate_extent_size(mp, tp, minlen, maxlen, len,
				&sumbp,	&sb, prod, &r);
		break;
	case XFS_ALLOCTYPE_NEAR_BNO:
		error = xfs_rtallocate_extent_near(mp, tp, bno, minlen, maxlen,
				len, &sumbp, &sb, prod, &r);
		break;
	case XFS_ALLOCTYPE_THIS_BNO:
		error = xfs_rtallocate_extent_exact(mp, tp, bno, minlen, maxlen,
				len, &sumbp, &sb, prod, &r);
		break;
	default:
		ASSERT(0);
	}
	if (error) {
		return error;
	}
	/*
	 * If it worked, update the superblock.
	 */
	if (r != NULLRTBLOCK) {
		long	slen = (long)*len;

		ASSERT(*len >= minlen && *len <= maxlen);
		if (wasdel)
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_RES_FREXTENTS, -slen);
		else
			xfs_trans_mod_sb(tp, XFS_TRANS_SB_FREXTENTS, -slen);
	}
	*rtblock = r;
	return 0;
}

/*
 * Free an extent in the realtime subvolume.  Length is expressed in
 * realtime extents, as is the block number.
 */
int					/* error */
xfs_rtfree_extent(
	xfs_trans_t	*tp,		/* transaction pointer */
	xfs_rtblock_t	bno,		/* starting block number to free */
	xfs_extlen_t	len)		/* length of extent freed */
{
	int		error;		/* error value */
	xfs_inode_t	*ip;		/* bitmap file inode */
	xfs_mount_t	*mp;		/* file system mount structure */
	xfs_fsblock_t	sb;		/* summary file block number */
	xfs_buf_t	*sumbp;		/* summary file block buffer */

	mp = tp->t_mountp;
	/*
	 * Synchronize by locking the bitmap inode.
	 */
	if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0,
					XFS_ILOCK_EXCL, &ip)))
		return error;
#if defined(__KERNEL__) && defined(DEBUG)
	/*
	 * Check to see that this whole range is currently allocated.
	 */
	{
		int	stat;		/* result from checking range */

		error = xfs_rtcheck_alloc_range(mp, tp, bno, len, &stat);
		if (error) {
			return error;
		}
		ASSERT(stat);
	}
#endif
	sumbp = NULL;
	/*
	 * Free the range of realtime blocks.
	 */
	error = xfs_rtfree_range(mp, tp, bno, len, &sumbp, &sb);
	if (error) {
		return error;
	}
	/*
	 * Mark more blocks free in the superblock.
	 */
	xfs_trans_mod_sb(tp, XFS_TRANS_SB_FREXTENTS, (long)len);
	/*
	 * If we've now freed all the blocks, reset the file sequence
	 * number to 0.
	 */
	if (tp->t_frextents_delta + mp->m_sb.sb_frextents ==
	    mp->m_sb.sb_rextents) {
		if (!(ip->i_d.di_flags & XFS_DIFLAG_NEWRTBM))
			ip->i_d.di_flags |= XFS_DIFLAG_NEWRTBM;
		*(__uint64_t *)&ip->i_d.di_atime = 0;
		xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
	}
	return 0;
}

/*
 * Initialize realtime fields in the mount structure.
 */
int				/* error */
xfs_rtmount_init(
	xfs_mount_t	*mp)	/* file system mount structure */
{
	xfs_buf_t	*bp;	/* buffer for last block of subvolume */
	xfs_daddr_t	d;	/* address of last block of subvolume */
	int		error;	/* error return value */
	xfs_sb_t	*sbp;	/* filesystem superblock copy in mount */

	sbp = &mp->m_sb;
	if (sbp->sb_rblocks == 0)
		return 0;
	if (mp->m_rtdev_targp == NULL) {
		cmn_err(CE_WARN,
	"XFS: This filesystem has a realtime volume, use rtdev=device option");
		return XFS_ERROR(ENODEV);
	}
	mp->m_rsumlevels = sbp->sb_rextslog + 1;
	mp->m_rsumsize =
		(uint)sizeof(xfs_suminfo_t) * mp->m_rsumlevels *
		sbp->sb_rbmblocks;
	mp->m_rsumsize = roundup(mp->m_rsumsize, sbp->sb_blocksize);
	mp->m_rbmip = mp->m_rsumip = NULL;
	/*
	 * Check that the realtime section is an ok size.
	 */
	d = (xfs_daddr_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_rblocks);
	if (XFS_BB_TO_FSB(mp, d) != mp->m_sb.sb_rblocks) {
		cmn_err(CE_WARN, "XFS: realtime mount -- %llu != %llu",
			(unsigned long long) XFS_BB_TO_FSB(mp, d),
			(unsigned long long) mp->m_sb.sb_rblocks);
		return XFS_ERROR(E2BIG);
	}
	error = xfs_read_buf(mp, mp->m_rtdev_targp,
				d - XFS_FSB_TO_BB(mp, 1),
				XFS_FSB_TO_BB(mp, 1), 0, &bp);
	if (error) {
		cmn_err(CE_WARN,
	"XFS: realtime mount -- xfs_read_buf failed, returned %d", error);
		if (error == ENOSPC)
			return XFS_ERROR(E2BIG);
		return error;
	}
	xfs_buf_relse(bp);
	return 0;
}

/*
 * Get the bitmap and summary inodes into the mount structure
 * at mount time.
 */
int					/* error */
xfs_rtmount_inodes(
	xfs_mount_t	*mp)		/* file system mount structure */
{
	int		error;		/* error return value */
	xfs_sb_t	*sbp;

	sbp = &mp->m_sb;
	if (sbp->sb_rbmino == NULLFSINO)
		return 0;
	error = xfs_iget(mp, NULL, sbp->sb_rbmino, 0, 0, &mp->m_rbmip, 0);
	if (error)
		return error;
	ASSERT(mp->m_rbmip != NULL);
	ASSERT(sbp->sb_rsumino != NULLFSINO);
	error = xfs_iget(mp, NULL, sbp->sb_rsumino, 0, 0, &mp->m_rsumip, 0);
	if (error) {
		VN_RELE(XFS_ITOV(mp->m_rbmip));
		return error;
	}
	ASSERT(mp->m_rsumip != NULL);
	return 0;
}

/*
 * Pick an extent for allocation at the start of a new realtime file.
 * Use the sequence number stored in the atime field of the bitmap inode.
 * Translate this to a fraction of the rtextents, and return the product
 * of rtextents and the fraction.
 * The fraction sequence is 0, 1/2, 1/4, 3/4, 1/8, ..., 7/8, 1/16, ...
 */
int					/* error */
xfs_rtpick_extent(
	xfs_mount_t	*mp,		/* file system mount point */
	xfs_trans_t	*tp,		/* transaction pointer */
	xfs_extlen_t	len,		/* allocation length (rtextents) */
	xfs_rtblock_t	*pick)		/* result rt extent */
{
	xfs_rtblock_t	b;		/* result block */
	int		error;		/* error return value */
	xfs_inode_t	*ip;		/* bitmap incore inode */
	int		log2;		/* log of sequence number */
	__uint64_t	resid;		/* residual after log removed */
	__uint64_t	seq;		/* sequence number of file creation */
	__uint64_t	*seqp;		/* pointer to seqno in inode */

	if ((error = xfs_trans_iget(mp, tp, mp->m_sb.sb_rbmino, 0,
					XFS_ILOCK_EXCL, &ip)))
		return error;
	ASSERT(ip == mp->m_rbmip);
	seqp = (__uint64_t *)&ip->i_d.di_atime;
	if (!(ip->i_d.di_flags & XFS_DIFLAG_NEWRTBM)) {
		ip->i_d.di_flags |= XFS_DIFLAG_NEWRTBM;
		*seqp = 0;
	}
	seq = *seqp;
	if ((log2 = xfs_highbit64(seq)) == -1)
		b = 0;
	else {
		resid = seq - (1ULL << log2);
		b = (mp->m_sb.sb_rextents * ((resid << 1) + 1ULL)) >>
		    (log2 + 1);
		if (b >= mp->m_sb.sb_rextents)
			b = do_mod(b, mp->m_sb.sb_rextents);
		if (b + len > mp->m_sb.sb_rextents)
			b = mp->m_sb.sb_rextents - len;
	}
	*seqp = seq + 1;
	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
	*pick = b;
	return 0;
}