xfs_log_recover.c

linux 内核源代码

/*
 * Copyright (c) 2000-2006 Silicon Graphics, 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_dir2.h"
#include "xfs_dmapi.h"
#include "xfs_mount.h"
#include "xfs_error.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
#include "xfs_dir2_sf.h"
#include "xfs_attr_sf.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_inode_item.h"
#include "xfs_imap.h"
#include "xfs_alloc.h"
#include "xfs_ialloc.h"
#include "xfs_log_priv.h"
#include "xfs_buf_item.h"
#include "xfs_log_recover.h"
#include "xfs_extfree_item.h"
#include "xfs_trans_priv.h"
#include "xfs_quota.h"
#include "xfs_rw.h"

STATIC int	xlog_find_zeroed(xlog_t *, xfs_daddr_t *);
STATIC int	xlog_clear_stale_blocks(xlog_t *, xfs_lsn_t);
STATIC void	xlog_recover_insert_item_backq(xlog_recover_item_t **q,
					       xlog_recover_item_t *item);
#if defined(DEBUG)
STATIC void	xlog_recover_check_summary(xlog_t *);
STATIC void	xlog_recover_check_ail(xfs_mount_t *, xfs_log_item_t *, int);
#else
#define	xlog_recover_check_summary(log)
#define	xlog_recover_check_ail(mp, lip, gen)
#endif


/*
 * Sector aligned buffer routines for buffer create/read/write/access
 */

#define XLOG_SECTOR_ROUNDUP_BBCOUNT(log, bbs)	\
	( ((log)->l_sectbb_mask && (bbs & (log)->l_sectbb_mask)) ? \
	((bbs + (log)->l_sectbb_mask + 1) & ~(log)->l_sectbb_mask) : (bbs) )
#define XLOG_SECTOR_ROUNDDOWN_BLKNO(log, bno)	((bno) & ~(log)->l_sectbb_mask)

xfs_buf_t *
xlog_get_bp(
	xlog_t		*log,
	int		num_bblks)
{
	ASSERT(num_bblks > 0);

	if (log->l_sectbb_log) {
		if (num_bblks > 1)
			num_bblks += XLOG_SECTOR_ROUNDUP_BBCOUNT(log, 1);
		num_bblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, num_bblks);
	}
	return xfs_buf_get_noaddr(BBTOB(num_bblks), log->l_mp->m_logdev_targp);
}

void
xlog_put_bp(
	xfs_buf_t	*bp)
{
	xfs_buf_free(bp);
}


/*
 * nbblks should be uint, but oh well.  Just want to catch that 32-bit length.
 */
int
xlog_bread(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	int		error;

	if (log->l_sectbb_log) {
		blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no);
		nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks);
	}

	ASSERT(nbblks > 0);
	ASSERT(BBTOB(nbblks) <= XFS_BUF_SIZE(bp));
	ASSERT(bp);

	XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
	XFS_BUF_READ(bp);
	XFS_BUF_BUSY(bp);
	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
	XFS_BUF_SET_TARGET(bp, log->l_mp->m_logdev_targp);

	xfsbdstrat(log->l_mp, bp);
	if ((error = xfs_iowait(bp)))
		xfs_ioerror_alert("xlog_bread", log->l_mp,
				  bp, XFS_BUF_ADDR(bp));
	return error;
}

/*
 * Write out the buffer at the given block for the given number of blocks.
 * The buffer is kept locked across the write and is returned locked.
 * This can only be used for synchronous log writes.
 */
STATIC int
xlog_bwrite(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	int		error;

	if (log->l_sectbb_log) {
		blk_no = XLOG_SECTOR_ROUNDDOWN_BLKNO(log, blk_no);
		nbblks = XLOG_SECTOR_ROUNDUP_BBCOUNT(log, nbblks);
	}

	ASSERT(nbblks > 0);
	ASSERT(BBTOB(nbblks) <= XFS_BUF_SIZE(bp));

	XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
	XFS_BUF_ZEROFLAGS(bp);
	XFS_BUF_BUSY(bp);
	XFS_BUF_HOLD(bp);
	XFS_BUF_PSEMA(bp, PRIBIO);
	XFS_BUF_SET_COUNT(bp, BBTOB(nbblks));
	XFS_BUF_SET_TARGET(bp, log->l_mp->m_logdev_targp);

	if ((error = xfs_bwrite(log->l_mp, bp)))
		xfs_ioerror_alert("xlog_bwrite", log->l_mp,
				  bp, XFS_BUF_ADDR(bp));
	return error;
}

STATIC xfs_caddr_t
xlog_align(
	xlog_t		*log,
	xfs_daddr_t	blk_no,
	int		nbblks,
	xfs_buf_t	*bp)
{
	xfs_caddr_t	ptr;

	if (!log->l_sectbb_log)
		return XFS_BUF_PTR(bp);

	ptr = XFS_BUF_PTR(bp) + BBTOB((int)blk_no & log->l_sectbb_mask);
	ASSERT(XFS_BUF_SIZE(bp) >=
		BBTOB(nbblks + (blk_no & log->l_sectbb_mask)));
	return ptr;
}

#ifdef DEBUG
/*
 * dump debug superblock and log record information
 */
STATIC void
xlog_header_check_dump(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	int			b;

	cmn_err(CE_DEBUG, "%s:  SB : uuid = ", __FUNCTION__);
	for (b = 0; b < 16; b++)
		cmn_err(CE_DEBUG, "%02x", ((uchar_t *)&mp->m_sb.sb_uuid)[b]);
	cmn_err(CE_DEBUG, ", fmt = %d\n", XLOG_FMT);
	cmn_err(CE_DEBUG, "    log : uuid = ");
	for (b = 0; b < 16; b++)
		cmn_err(CE_DEBUG, "%02x",((uchar_t *)&head->h_fs_uuid)[b]);
	cmn_err(CE_DEBUG, ", fmt = %d\n", INT_GET(head->h_fmt, ARCH_CONVERT));
}
#else
#define xlog_header_check_dump(mp, head)
#endif

/*
 * check log record header for recovery
 */
STATIC int
xlog_header_check_recover(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	ASSERT(INT_GET(head->h_magicno, ARCH_CONVERT) == XLOG_HEADER_MAGIC_NUM);

	/*
	 * IRIX doesn't write the h_fmt field and leaves it zeroed
	 * (XLOG_FMT_UNKNOWN). This stops us from trying to recover
	 * a dirty log created in IRIX.
	 */
	if (unlikely(INT_GET(head->h_fmt, ARCH_CONVERT) != XLOG_FMT)) {
		xlog_warn(
	"XFS: dirty log written in incompatible format - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_recover(1)",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
		xlog_warn(
	"XFS: dirty log entry has mismatched uuid - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_recover(2)",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	return 0;
}

/*
 * read the head block of the log and check the header
 */
STATIC int
xlog_header_check_mount(
	xfs_mount_t		*mp,
	xlog_rec_header_t	*head)
{
	ASSERT(INT_GET(head->h_magicno, ARCH_CONVERT) == XLOG_HEADER_MAGIC_NUM);

	if (uuid_is_nil(&head->h_fs_uuid)) {
		/*
		 * IRIX doesn't write the h_fs_uuid or h_fmt fields. If
		 * h_fs_uuid is nil, we assume this log was last mounted
		 * by IRIX and continue.
		 */
		xlog_warn("XFS: nil uuid in log - IRIX style log");
	} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
		xlog_warn("XFS: log has mismatched uuid - can't recover");
		xlog_header_check_dump(mp, head);
		XFS_ERROR_REPORT("xlog_header_check_mount",
				 XFS_ERRLEVEL_HIGH, mp);
		return XFS_ERROR(EFSCORRUPTED);
	}
	return 0;
}

STATIC void
xlog_recover_iodone(
	struct xfs_buf	*bp)
{
	xfs_mount_t	*mp;

	ASSERT(XFS_BUF_FSPRIVATE(bp, void *));

	if (XFS_BUF_GETERROR(bp)) {
		/*
		 * We're not going to bother about retrying
		 * this during recovery. One strike!
		 */
		mp = XFS_BUF_FSPRIVATE(bp, xfs_mount_t *);
		xfs_ioerror_alert("xlog_recover_iodone",
				  mp, bp, XFS_BUF_ADDR(bp));
		xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR);
	}
	XFS_BUF_SET_FSPRIVATE(bp, NULL);
	XFS_BUF_CLR_IODONE_FUNC(bp);
	xfs_biodone(bp);
}

/*
 * This routine finds (to an approximation) the first block in the physical
 * log which contains the given cycle.  It uses a binary search algorithm.
 * Note that the algorithm can not be perfect because the disk will not
 * necessarily be perfect.
 */
int
xlog_find_cycle_start(
	xlog_t		*log,
	xfs_buf_t	*bp,
	xfs_daddr_t	first_blk,
	xfs_daddr_t	*last_blk,
	uint		cycle)
{
	xfs_caddr_t	offset;
	xfs_daddr_t	mid_blk;
	uint		mid_cycle;
	int		error;

	mid_blk = BLK_AVG(first_blk, *last_blk);
	while (mid_blk != first_blk && mid_blk != *last_blk) {
		if ((error = xlog_bread(log, mid_blk, 1, bp)))
			return error;
		offset = xlog_align(log, mid_blk, 1, bp);
		mid_cycle = GET_CYCLE(offset, ARCH_CONVERT);
		if (mid_cycle == cycle) {
			*last_blk = mid_blk;
			/* last_half_cycle == mid_cycle */
		} else {
			first_blk = mid_blk;
			/* first_half_cycle == mid_cycle */
		}
		mid_blk = BLK_AVG(first_blk, *last_blk);
	}
	ASSERT((mid_blk == first_blk && mid_blk+1 == *last_blk) ||
	       (mid_blk == *last_blk && mid_blk-1 == first_blk));

	return 0;
}

/*
 * Check that the range of blocks does not contain the cycle number
 * given.  The scan needs to occur from front to back and the ptr into the
 * region must be updated since a later routine will need to perform another
 * test.  If the region is completely good, we end up returning the same
 * last block number.
 *
 * Set blkno to -1 if we encounter no errors.  This is an invalid block number
 * since we don't ever expect logs to get this large.
 */
STATIC int
xlog_find_verify_cycle(
	xlog_t		*log,
	xfs_daddr_t	start_blk,
	int		nbblks,
	uint		stop_on_cycle_no,
	xfs_daddr_t	*new_blk)
{
	xfs_daddr_t	i, j;
	uint		cycle;
	xfs_buf_t	*bp;
	xfs_daddr_t	bufblks;
	xfs_caddr_t	buf = NULL;
	int		error = 0;

	bufblks = 1 << ffs(nbblks);

	while (!(bp = xlog_get_bp(log, bufblks))) {
		/* can't get enough memory to do everything in one big buffer */
		bufblks >>= 1;
		if (bufblks <= log->l_sectbb_log)
			return ENOMEM;
	}

	for (i = start_blk; i < start_blk + nbblks; i += bufblks) {
		int	bcount;

		bcount = min(bufblks, (start_blk + nbblks - i));

		if ((error = xlog_bread(log, i, bcount, bp)))
			goto out;

		buf = xlog_align(log, i, bcount, bp);
		for (j = 0; j < bcount; j++) {
			cycle = GET_CYCLE(buf, ARCH_CONVERT);
			if (cycle == stop_on_cycle_no) {
				*new_blk = i+j;
				goto out;
			}

			buf += BBSIZE;
		}
	}

	*new_blk = -1;

out:
	xlog_put_bp(bp);
	return error;
}

/*
 * Potentially backup over partial log record write.
 *
 * In the typical case, last_blk is the number of the block directly after
 * a good log record.  Therefore, we subtract one to get the block number
 * of the last block in the given buffer.  extra_bblks contains the number
 * of blocks we would have read on a previous read.  This happens when the
 * last log record is split over the end of the physical log.
 *
 * extra_bblks is the number of blocks potentially verified on a previous
 * call to this routine.
 */
STATIC int
xlog_find_verify_log_record(
	xlog_t			*log,
	xfs_daddr_t		start_blk,
	xfs_daddr_t		*last_blk,
	int			extra_bblks)
{
	xfs_daddr_t		i;
	xfs_buf_t		*bp;
	xfs_caddr_t		offset = NULL;
	xlog_rec_header_t	*head = NULL;
	int			error = 0;
	int			smallmem = 0;
	int			num_blks = *last_blk - start_blk;
	int			xhdrs;

	ASSERT(start_blk != 0 || *last_blk != start_blk);

	if (!(bp = xlog_get_bp(log, num_blks))) {
		if (!(bp = xlog_get_bp(log, 1)))
			return ENOMEM;
		smallmem = 1;
	} else {
		if ((error = xlog_bread(log, start_blk, num_blks, bp)))
			goto out;
		offset = xlog_align(log, start_blk, num_blks, bp);
		offset += ((num_blks - 1) << BBSHIFT);
	}

	for (i = (*last_blk) - 1; i >= 0; i--) {
		if (i < start_blk) {
			/* valid log record not found */
			xlog_warn(
		"XFS: Log inconsistent (didn't find previous header)");
			ASSERT(0);
			error = XFS_ERROR(EIO);
			goto out;
		}

		if (smallmem) {
			if ((error = xlog_bread(log, i, 1, bp)))
				goto out;
			offset = xlog_align(log, i, 1, bp);
		}

		head = (xlog_rec_header_t *)offset;

		if (XLOG_HEADER_MAGIC_NUM ==
		    INT_GET(head->h_magicno, ARCH_CONVERT))
			break;

		if (!smallmem)
			offset -= BBSIZE;
	}

	/*
	 * We hit the beginning of the physical log & still no header.  Return
	 * to caller.  If caller can handle a return of -1, then this routine
	 * will be called again for the end of the physical log.
	 */
	if (i == -1) {