path: root/fs/xfs/xfs_trans.c

/*
 * Copyright (c) 2000-2003,2005 Silicon Graphics, Inc.
 * Copyright (C) 2010 Red Hat, Inc.
 * All Rights Reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it would be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include "xfs.h"
#include "xfs_fs.h"
#include "xfs_types.h"
#include "xfs_bit.h"
#include "xfs_log.h"
#include "xfs_inum.h"
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_mount.h"
#include "xfs_error.h"
#include "xfs_da_btree.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_btree.h"
#include "xfs_ialloc.h"
#include "xfs_alloc.h"
#include "xfs_bmap.h"
#include "xfs_quota.h"
#include "xfs_trans_priv.h"
#include "xfs_trans_space.h"
#include "xfs_inode_item.h"
#include "xfs_trace.h"

kmem_zone_t	*xfs_trans_zone;
kmem_zone_t	*xfs_log_item_desc_zone;


/*
 * Various log reservation values.
 *
 * These are based on the size of the file system block because that is what
 * most transactions manipulate.  Each adds in an additional 128 bytes per
 * item logged to try to account for the overhead of the transaction mechanism.
 *
 * Note:  Most of the reservations underestimate the number of allocation
 * groups into which they could free extents in the xfs_bmap_finish() call.
 * This is because the number in the worst case is quite high and quite
 * unusual.  In order to fix this we need to change xfs_bmap_finish() to free
 * extents in only a single AG at a time.  This will require changes to the
 * EFI code as well, however, so that the EFI for the extents not freed is
 * logged again in each transaction.  See SGI PV #261917.
 *
 * Reservation functions here avoid a huge stack in xfs_trans_init due to
 * register overflow from temporaries in the calculations.
 */


/*
 * In a write transaction we can allocate a maximum of 2
 * extents.  This gives:
 *    the inode getting the new extents: inode size
 *    the inode's bmap btree: max depth * block size
 *    the agfs of the ags from which the extents are allocated: 2 * sector
 *    the superblock free block counter: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 * And the bmap_finish transaction can free bmap blocks in a join:
 *    the agfs of the ags containing the blocks: 2 * sector size
 *    the agfls of the ags containing the blocks: 2 * sector size
 *    the super block free block counter: sector size
 *    the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_write_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (4 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 2))),
		    (2 * mp->m_sb.sb_sectsize +
		     2 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 2) +
		     128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2))));
}

/*
 * In truncating a file we free up to two extents at once.  We can modify:
 *    the inode being truncated: inode size
 *    the inode's bmap btree: (max depth + 1) * block size
 * And the bmap_finish transaction can free the blocks and bmap blocks:
 *    the agf for each of the ags: 4 * sector size
 *    the agfl for each of the ags: 4 * sector size
 *    the super block to reflect the freed blocks: sector size
 *    worst case split in allocation btrees per extent assuming 4 extents:
 *		4 exts * 2 trees * (2 * max depth - 1) * block size
 *    the inode btree: max depth * blocksize
 *    the allocation btrees: 2 trees * (max depth - 1) * block size
 */
STATIC uint
xfs_calc_itruncate_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1) +
		     128 * (2 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))),
		    (4 * mp->m_sb.sb_sectsize +
		     4 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 4) +
		     128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)) +
		     128 * 5 +
		     XFS_ALLOCFREE_LOG_RES(mp, 1) +
		     128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels +
			    XFS_ALLOCFREE_LOG_COUNT(mp, 1))));
}

/*
 * In renaming a files we can modify:
 *    the four inodes involved: 4 * inode size
 *    the two directory btrees: 2 * (max depth + v2) * dir block size
 *    the two directory bmap btrees: 2 * max depth * block size
 * And the bmap_finish transaction can free dir and bmap blocks (two sets
 *	of bmap blocks) giving:
 *    the agf for the ags in which the blocks live: 3 * sector size
 *    the agfl for the ags in which the blocks live: 3 * sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_rename_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((4 * mp->m_sb.sb_inodesize +
		     2 * XFS_DIROP_LOG_RES(mp) +
		     128 * (4 + 2 * XFS_DIROP_LOG_COUNT(mp))),
		    (3 * mp->m_sb.sb_sectsize +
		     3 * mp->m_sb.sb_sectsize +
		     mp->m_sb.sb_sectsize +
		     XFS_ALLOCFREE_LOG_RES(mp, 3) +
		     128 * (7 + XFS_ALLOCFREE_LOG_COUNT(mp, 3))));
}

/*
 * For creating a link to an inode:
 *    the parent directory inode: inode size
 *    the linked inode: inode size
 *    the directory btree could split: (max depth + v2) * dir block size
 *    the directory bmap btree could join or split: (max depth + v2) * blocksize
 * And the bmap_finish transaction can free some bmap blocks giving:
 *    the agf for the ag in which the blocks live: sector size
 *    the agfl for the ag in which the blocks live: sector size
 *    the superblock for the free block count: sector size
 *    the allocation btrees: 2 trees * (2 * max depth - 1) * block size
 */
STATIC uint
xfs_calc_link_reservation(
	struct xfs_mount	*mp)
{
	return XFS_DQUOT_LOGRES(mp) +
		MAX((mp->m_sb.sb_inodesize +
		     mp->m_sb.sb_inodesize +
		     XFS_DIROP_LOG_RES(mp) +
		     128 * (2 + XFS_DIROP_LOG_