xfs_iomap.c

linux 内核源代码

		resrtextents /= mp->m_sb.sb_rextsize;
		resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0);
		quota_flag = XFS_QMOPT_RES_RTBLKS;
	} else {
		resrtextents = 0;
		resblks = qblocks = XFS_DIOSTRAT_SPACE_RES(mp, resaligned);
		quota_flag = XFS_QMOPT_RES_REGBLKS;
	}

	/*
	 * Allocate and setup the transaction
	 */
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	tp = xfs_trans_alloc(mp, XFS_TRANS_DIOSTRAT);
	error = xfs_trans_reserve(tp, resblks,
			XFS_WRITE_LOG_RES(mp), resrtextents,
			XFS_TRANS_PERM_LOG_RES,
			XFS_WRITE_LOG_COUNT);
	/*
	 * Check for running out of space, note: need lock to return
	 */
	if (error)
		xfs_trans_cancel(tp, 0);
	xfs_ilock(ip, XFS_ILOCK_EXCL);
	if (error)
		goto error_out;

	error = XFS_TRANS_RESERVE_QUOTA_NBLKS(mp, tp, ip,
					      qblocks, 0, quota_flag);
	if (error)
		goto error1;

	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
	xfs_trans_ihold(tp, ip);

	bmapi_flag = XFS_BMAPI_WRITE;
	if ((flags & BMAPI_DIRECT) && (offset < ip->i_size || extsz))
		bmapi_flag |= XFS_BMAPI_PREALLOC;

	/*
	 * Issue the xfs_bmapi() call to allocate the blocks
	 */
	XFS_BMAP_INIT(&free_list, &firstfsb);
	nimaps = 1;
	error = XFS_BMAPI(mp, tp, io, offset_fsb, count_fsb, bmapi_flag,
		&firstfsb, 0, &imap, &nimaps, &free_list, NULL);
	if (error)
		goto error0;

	/*
	 * Complete the transaction
	 */
	error = xfs_bmap_finish(&tp, &free_list, &committed);
	if (error)
		goto error0;
	error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
	if (error)
		goto error_out;

	/*
	 * Copy any maps to caller's array and return any error.
	 */
	if (nimaps == 0) {
		error = ENOSPC;
		goto error_out;
	}

	if (unlikely(!imap.br_startblock && !(io->io_flags & XFS_IOCORE_RT))) {
		error = xfs_cmn_err_fsblock_zero(ip, &imap);
		goto error_out;
	}

	*ret_imap = imap;
	*nmaps = 1;
	return 0;

error0:	/* Cancel bmap, unlock inode, unreserve quota blocks, cancel trans */
	xfs_bmap_cancel(&free_list);
	XFS_TRANS_UNRESERVE_QUOTA_NBLKS(mp, tp, ip, qblocks, 0, quota_flag);

error1:	/* Just cancel transaction */
	xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
	*nmaps = 0;	/* nothing set-up here */

error_out:
	return XFS_ERROR(error);
}

/*
 * If the caller is doing a write at the end of the file,
 * then extend the allocation out to the file system's write
 * iosize.  We clean up any extra space left over when the
 * file is closed in xfs_inactive().
 *
 * For sync writes, we are flushing delayed allocate space to
 * try to make additional space available for allocation near
 * the filesystem full boundary - preallocation hurts in that
 * situation, of course.
 */
STATIC int
xfs_iomap_eof_want_preallocate(
	xfs_mount_t	*mp,
	xfs_iocore_t	*io,
	xfs_fsize_t	isize,
	xfs_off_t	offset,
	size_t		count,
	int		ioflag,
	xfs_bmbt_irec_t *imap,
	int		nimaps,
	int		*prealloc)
{
	xfs_fileoff_t   start_fsb;
	xfs_filblks_t   count_fsb;
	xfs_fsblock_t	firstblock;
	int		n, error, imaps;

	*prealloc = 0;
	if ((ioflag & BMAPI_SYNC) || (offset + count) <= isize)
		return 0;

	/*
	 * If there are any real blocks past eof, then don't
	 * do any speculative allocation.
	 */
	start_fsb = XFS_B_TO_FSBT(mp, ((xfs_ufsize_t)(offset + count - 1)));
	count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_MAXIOFFSET(mp));
	while (count_fsb > 0) {
		imaps = nimaps;
		firstblock = NULLFSBLOCK;
		error = XFS_BMAPI(mp, NULL, io, start_fsb, count_fsb, 0,
				  &firstblock, 0, imap, &imaps, NULL, NULL);
		if (error)
			return error;
		for (n = 0; n < imaps; n++) {
			if ((imap[n].br_startblock != HOLESTARTBLOCK) &&
			    (imap[n].br_startblock != DELAYSTARTBLOCK))
				return 0;
			start_fsb += imap[n].br_blockcount;
			count_fsb -= imap[n].br_blockcount;
		}
	}
	*prealloc = 1;
	return 0;
}

int
xfs_iomap_write_delay(
	xfs_inode_t	*ip,
	xfs_off_t	offset,
	size_t		count,
	int		ioflag,
	xfs_bmbt_irec_t *ret_imap,
	int		*nmaps)
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_iocore_t	*io = &ip->i_iocore;
	xfs_fileoff_t	offset_fsb;
	xfs_fileoff_t	last_fsb;
	xfs_off_t	aligned_offset;
	xfs_fileoff_t	ioalign;
	xfs_fsblock_t	firstblock;
	xfs_extlen_t	extsz;
	xfs_fsize_t	isize;
	int		nimaps;
	xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS];
	int		prealloc, fsynced = 0;
	int		error;

	ASSERT(ismrlocked(&ip->i_lock, MR_UPDATE) != 0);

	/*
	 * Make sure that the dquots are there. This doesn't hold
	 * the ilock across a disk read.
	 */
	error = XFS_QM_DQATTACH(mp, ip, XFS_QMOPT_ILOCKED);
	if (error)
		return XFS_ERROR(error);

	extsz = xfs_get_extsz_hint(ip);
	offset_fsb = XFS_B_TO_FSBT(mp, offset);

retry:
	isize = ip->i_size;
	if (io->io_new_size > isize)
		isize = io->io_new_size;

	error = xfs_iomap_eof_want_preallocate(mp, io, isize, offset, count,
				ioflag, imap, XFS_WRITE_IMAPS, &prealloc);
	if (error)
		return error;

	if (prealloc) {
		aligned_offset = XFS_WRITEIO_ALIGN(mp, (offset + count - 1));
		ioalign = XFS_B_TO_FSBT(mp, aligned_offset);
		last_fsb = ioalign + mp->m_writeio_blocks;
	} else {
		last_fsb = XFS_B_TO_FSB(mp, ((xfs_ufsize_t)(offset + count)));
	}

	if (prealloc || extsz) {
		error = xfs_iomap_eof_align_last_fsb(mp, io, isize, extsz,
							&last_fsb);
		if (error)
			return error;
	}

	nimaps = XFS_WRITE_IMAPS;
	firstblock = NULLFSBLOCK;
	error = XFS_BMAPI(mp, NULL, io, offset_fsb,
			  (xfs_filblks_t)(last_fsb - offset_fsb),
			  XFS_BMAPI_DELAY | XFS_BMAPI_WRITE |
			  XFS_BMAPI_ENTIRE, &firstblock, 1, imap,
			  &nimaps, NULL, NULL);
	if (error && (error != ENOSPC))
		return XFS_ERROR(error);

	/*
	 * If bmapi returned us nothing, and if we didn't get back EDQUOT,
	 * then we must have run out of space - flush delalloc, and retry..
	 */
	if (nimaps == 0) {
		xfs_iomap_enter_trace(XFS_IOMAP_WRITE_NOSPACE,
					io, offset, count);
		if (xfs_flush_space(ip, &fsynced, &ioflag))
			return XFS_ERROR(ENOSPC);

		error = 0;
		goto retry;
	}

	if (unlikely(!imap[0].br_startblock && !(io->io_flags & XFS_IOCORE_RT)))
		return xfs_cmn_err_fsblock_zero(ip, &imap[0]);

	*ret_imap = imap[0];
	*nmaps = 1;

	return 0;
}

/*
 * Pass in a delayed allocate extent, convert it to real extents;
 * return to the caller the extent we create which maps on top of
 * the originating callers request.
 *
 * Called without a lock on the inode.
 */
int
xfs_iomap_write_allocate(
	xfs_inode_t	*ip,
	xfs_off_t	offset,
	size_t		count,
	xfs_bmbt_irec_t *map,
	int		*retmap)
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_iocore_t    *io = &ip->i_iocore;
	xfs_fileoff_t	offset_fsb, last_block;
	xfs_fileoff_t	end_fsb, map_start_fsb;
	xfs_fsblock_t	first_block;
	xfs_bmap_free_t	free_list;
	xfs_filblks_t	count_fsb;
	xfs_bmbt_irec_t	imap[XFS_STRAT_WRITE_IMAPS];
	xfs_trans_t	*tp;
	int		i, nimaps, committed;
	int		error = 0;
	int		nres;

	*retmap = 0;

	/*
	 * Make sure that the dquots are there.
	 */
	if ((error = XFS_QM_DQATTACH(mp, ip, 0)))
		return XFS_ERROR(error);

	offset_fsb = XFS_B_TO_FSBT(mp, offset);
	count_fsb = map->br_blockcount;
	map_start_fsb = map->br_startoff;

	XFS_STATS_ADD(xs_xstrat_bytes, XFS_FSB_TO_B(mp, count_fsb));

	while (count_fsb != 0) {
		/*
		 * Set up a transaction with which to allocate the
		 * backing store for the file.  Do allocations in a
		 * loop until we get some space in the range we are
		 * interested in.  The other space that might be allocated
		 * is in the delayed allocation extent on which we sit
		 * but before our buffer starts.
		 */

		nimaps = 0;
		while (nimaps == 0) {
			tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
			tp->t_flags |= XFS_TRANS_RESERVE;
			nres = XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK);
			error = xfs_trans_reserve(tp, nres,
					XFS_WRITE_LOG_RES(mp),
					0, XFS_TRANS_PERM_LOG_RES,
					XFS_WRITE_LOG_COUNT);
			if (error) {
				xfs_trans_cancel(tp, 0);
				return XFS_ERROR(error);
			}
			xfs_ilock(ip, XFS_ILOCK_EXCL);
			xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
			xfs_trans_ihold(tp, ip);

			XFS_BMAP_INIT(&free_list, &first_block);

			nimaps = XFS_STRAT_WRITE_IMAPS;
			/*
			 * Ensure we don't go beyond eof - it is possible
			 * the extents changed since we did the read call,
			 * we dropped the ilock in the interim.
			 */

			end_fsb = XFS_B_TO_FSB(mp, ip->i_size);
			xfs_bmap_last_offset(NULL, ip, &last_block,
				XFS_DATA_FORK);
			last_block = XFS_FILEOFF_MAX(last_block, end_fsb);
			if ((map_start_fsb + count_fsb) > last_block) {
				count_fsb = last_block - map_start_fsb;
				if (count_fsb == 0) {
					error = EAGAIN;
					goto trans_cancel;
				}
			}

			/* Go get the actual blocks */
			error = XFS_BMAPI(mp, tp, io, map_start_fsb, count_fsb,
					XFS_BMAPI_WRITE, &first_block, 1,
					imap, &nimaps, &free_list, NULL);
			if (error)
				goto trans_cancel;

			error = xfs_bmap_finish(&tp, &free_list, &committed);
			if (error)
				goto trans_cancel;

			error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
			if (error)
				goto error0;

			xfs_iunlock(ip, XFS_ILOCK_EXCL);
		}

		/*
		 * See if we were able to allocate an extent that
		 * covers at least part of the callers request
		 */
		for (i = 0; i < nimaps; i++) {
			if (unlikely(!imap[i].br_startblock &&
				     !(io->io_flags & XFS_IOCORE_RT)))
				return xfs_cmn_err_fsblock_zero(ip, &imap[i]);
			if ((offset_fsb >= imap[i].br_startoff) &&
			    (offset_fsb < (imap[i].br_startoff +
					   imap[i].br_blockcount))) {
				*map = imap[i];
				*retmap = 1;
				XFS_STATS_INC(xs_xstrat_quick);
				return 0;
			}
			count_fsb -= imap[i].br_blockcount;
		}

		/* So far we have not mapped the requested part of the
		 * file, just surrounding data, try again.
		 */
		nimaps--;
		map_start_fsb = imap[nimaps].br_startoff +
				imap[nimaps].br_blockcount;
	}

trans_cancel:
	xfs_bmap_cancel(&free_list);
	xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT);
error0:
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	return XFS_ERROR(error);
}

int
xfs_iomap_write_unwritten(
	xfs_inode_t	*ip,
	xfs_off_t	offset,
	size_t		count)
{
	xfs_mount_t	*mp = ip->i_mount;
	xfs_iocore_t    *io = &ip->i_iocore;
	xfs_fileoff_t	offset_fsb;
	xfs_filblks_t	count_fsb;
	xfs_filblks_t	numblks_fsb;
	xfs_fsblock_t	firstfsb;
	int		nimaps;
	xfs_trans_t	*tp;
	xfs_bmbt_irec_t imap;
	xfs_bmap_free_t free_list;
	uint		resblks;
	int		committed;
	int		error;

	xfs_iomap_enter_trace(XFS_IOMAP_UNWRITTEN,
				&ip->i_iocore, offset, count);

	offset_fsb = XFS_B_TO_FSBT(mp, offset);
	count_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + count);
	count_fsb = (xfs_filblks_t)(count_fsb - offset_fsb);

	resblks = XFS_DIOSTRAT_SPACE_RES(mp, 0) << 1;

	do {
		/*
		 * set up a transaction to convert the range of extents
		 * from unwritten to real. Do allocations in a loop until
		 * we have covered the range passed in.
		 */
		tp = xfs_trans_alloc(mp, XFS_TRANS_STRAT_WRITE);
		tp->t_flags |= XFS_TRANS_RESERVE;
		error = xfs_trans_reserve(tp, resblks,
				XFS_WRITE_LOG_RES(mp), 0,
				XFS_TRANS_PERM_LOG_RES,
				XFS_WRITE_LOG_COUNT);
		if (error) {
			xfs_trans_cancel(tp, 0);
			return XFS_ERROR(error);
		}

		xfs_ilock(ip, XFS_ILOCK_EXCL);
		xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
		xfs_trans_ihold(tp, ip);

		/*
		 * Modify the unwritten extent state of the buffer.
		 */
		XFS_BMAP_INIT(&free_list, &firstfsb);
		nimaps = 1;
		error = XFS_BMAPI(mp, tp, io, offset_fsb, count_fsb,
				  XFS_BMAPI_WRITE|XFS_BMAPI_CONVERT, &firstfsb,
				  1, &imap, &nimaps, &free_list, NULL);
		if (error)
			goto error_on_bmapi_transaction;

		error = xfs_bmap_finish(&(tp), &(free_list), &committed);
		if (error)
			goto error_on_bmapi_transaction;

		error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
		xfs_iunlock(ip, XFS_ILOCK_EXCL);
		if (error)
			return XFS_ERROR(error);

		if (unlikely(!imap.br_startblock &&
			     !(io->io_flags & XFS_IOCORE_RT)))
			return xfs_cmn_err_fsblock_zero(ip, &imap);

		if ((numblks_fsb = imap.br_blockcount) == 0) {
			/*
			 * The numblks_fsb value should always get
			 * smaller, otherwise the loop is stuck.
			 */
			ASSERT(imap.br_blockcount);
			break;
		}
		offset_fsb += numblks_fsb;
		count_fsb -= numblks_fsb;
	} while (count_fsb > 0);

	return 0;

error_on_bmapi_transaction:
	xfs_bmap_cancel(&free_list);
	xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT));
	xfs_iunlock(ip, XFS_ILOCK_EXCL);
	return XFS_ERROR(error);
}