blob: cd0b077deb354b6de5dda287cffaeb351e0e8cce [file] [log] [blame]
/*
* 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_shared.h"
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_bit.h"
#include "xfs_sb.h"
#include "xfs_mount.h"
#include "xfs_da_format.h"
#include "xfs_da_btree.h"
#include "xfs_inode.h"
#include "xfs_trans.h"
#include "xfs_log.h"
#include "xfs_log_priv.h"
#include "xfs_log_recover.h"
#include "xfs_inode_item.h"
#include "xfs_extfree_item.h"
#include "xfs_trans_priv.h"
#include "xfs_alloc.h"
#include "xfs_ialloc.h"
#include "xfs_quota.h"
#include "xfs_cksum.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
#include "xfs_bmap_btree.h"
#include "xfs_error.h"
#include "xfs_dir2.h"
#include "xfs_rmap_item.h"
#include "xfs_buf_item.h"
#include "xfs_refcount_item.h"
#include "xfs_bmap_item.h"
#define BLK_AVG(blk1, blk2) ((blk1+blk2) >> 1)
STATIC int
xlog_find_zeroed(
struct xlog *,
xfs_daddr_t *);
STATIC int
xlog_clear_stale_blocks(
struct xlog *,
xfs_lsn_t);
#if defined(DEBUG)
STATIC void
xlog_recover_check_summary(
struct xlog *);
#else
#define xlog_recover_check_summary(log)
#endif
STATIC int
xlog_do_recovery_pass(
struct xlog *, xfs_daddr_t, xfs_daddr_t, int, xfs_daddr_t *);
/*
* This structure is used during recovery to record the buf log items which
* have been canceled and should not be replayed.
*/
struct xfs_buf_cancel {
xfs_daddr_t bc_blkno;
uint bc_len;
int bc_refcount;
struct list_head bc_list;
};
/*
* Sector aligned buffer routines for buffer create/read/write/access
*/
/*
* Verify the given count of basic blocks is valid number of blocks
* to specify for an operation involving the given XFS log buffer.
* Returns nonzero if the count is valid, 0 otherwise.
*/
static inline int
xlog_buf_bbcount_valid(
struct xlog *log,
int bbcount)
{
return bbcount > 0 && bbcount <= log->l_logBBsize;
}
/*
* Allocate a buffer to hold log data. The buffer needs to be able
* to map to a range of nbblks basic blocks at any valid (basic
* block) offset within the log.
*/
STATIC xfs_buf_t *
xlog_get_bp(
struct xlog *log,
int nbblks)
{
struct xfs_buf *bp;
if (!xlog_buf_bbcount_valid(log, nbblks)) {
xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer",
nbblks);
XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_HIGH, log->l_mp);
return NULL;
}
/*
* We do log I/O in units of log sectors (a power-of-2
* multiple of the basic block size), so we round up the
* requested size to accommodate the basic blocks required
* for complete log sectors.
*
* In addition, the buffer may be used for a non-sector-
* aligned block offset, in which case an I/O of the
* requested size could extend beyond the end of the
* buffer. If the requested size is only 1 basic block it
* will never straddle a sector boundary, so this won't be
* an issue. Nor will this be a problem if the log I/O is
* done in basic blocks (sector size 1). But otherwise we
* extend the buffer by one extra log sector to ensure
* there's space to accommodate this possibility.
*/
if (nbblks > 1 && log->l_sectBBsize > 1)
nbblks += log->l_sectBBsize;
nbblks = round_up(nbblks, log->l_sectBBsize);
bp = xfs_buf_get_uncached(log->l_mp->m_logdev_targp, nbblks, 0);
if (bp)
xfs_buf_unlock(bp);
return bp;
}
STATIC void
xlog_put_bp(
xfs_buf_t *bp)
{
xfs_buf_free(bp);
}
/*
* Return the address of the start of the given block number's data
* in a log buffer. The buffer covers a log sector-aligned region.
*/
STATIC char *
xlog_align(
struct xlog *log,
xfs_daddr_t blk_no,
int nbblks,
struct xfs_buf *bp)
{
xfs_daddr_t offset = blk_no & ((xfs_daddr_t)log->l_sectBBsize - 1);
ASSERT(offset + nbblks <= bp->b_length);
return bp->b_addr + BBTOB(offset);
}
/*
* nbblks should be uint, but oh well. Just want to catch that 32-bit length.
*/
STATIC int
xlog_bread_noalign(
struct xlog *log,
xfs_daddr_t blk_no,
int nbblks,
struct xfs_buf *bp)
{
int error;
if (!xlog_buf_bbcount_valid(log, nbblks)) {
xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer",
nbblks);
XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_HIGH, log->l_mp);
return -EFSCORRUPTED;
}
blk_no = round_down(blk_no, log->l_sectBBsize);
nbblks = round_up(nbblks, log->l_sectBBsize);
ASSERT(nbblks > 0);
ASSERT(nbblks <= bp->b_length);
XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
bp->b_flags |= XBF_READ;
bp->b_io_length = nbblks;
bp->b_error = 0;
error = xfs_buf_submit_wait(bp);
if (error && !XFS_FORCED_SHUTDOWN(log->l_mp))
xfs_buf_ioerror_alert(bp, __func__);
return error;
}
STATIC int
xlog_bread(
struct xlog *log,
xfs_daddr_t blk_no,
int nbblks,
struct xfs_buf *bp,
char **offset)
{
int error;
error = xlog_bread_noalign(log, blk_no, nbblks, bp);
if (error)
return error;
*offset = xlog_align(log, blk_no, nbblks, bp);
return 0;
}
/*
* Read at an offset into the buffer. Returns with the buffer in it's original
* state regardless of the result of the read.
*/
STATIC int
xlog_bread_offset(
struct xlog *log,
xfs_daddr_t blk_no, /* block to read from */
int nbblks, /* blocks to read */
struct xfs_buf *bp,
char *offset)
{
char *orig_offset = bp->b_addr;
int orig_len = BBTOB(bp->b_length);
int error, error2;
error = xfs_buf_associate_memory(bp, offset, BBTOB(nbblks));
if (error)
return error;
error = xlog_bread_noalign(log, blk_no, nbblks, bp);
/* must reset buffer pointer even on error */
error2 = xfs_buf_associate_memory(bp, orig_offset, orig_len);
if (error)
return error;
return error2;
}
/*
* 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(
struct xlog *log,
xfs_daddr_t blk_no,
int nbblks,
struct xfs_buf *bp)
{
int error;
if (!xlog_buf_bbcount_valid(log, nbblks)) {
xfs_warn(log->l_mp, "Invalid block length (0x%x) for buffer",
nbblks);
XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_HIGH, log->l_mp);
return -EFSCORRUPTED;
}
blk_no = round_down(blk_no, log->l_sectBBsize);
nbblks = round_up(nbblks, log->l_sectBBsize);
ASSERT(nbblks > 0);
ASSERT(nbblks <= bp->b_length);
XFS_BUF_SET_ADDR(bp, log->l_logBBstart + blk_no);
xfs_buf_hold(bp);
xfs_buf_lock(bp);
bp->b_io_length = nbblks;
bp->b_error = 0;
error = xfs_bwrite(bp);
if (error)
xfs_buf_ioerror_alert(bp, __func__);
xfs_buf_relse(bp);
return error;
}
#ifdef DEBUG
/*
* dump debug superblock and log record information
*/
STATIC void
xlog_header_check_dump(
xfs_mount_t *mp,
xlog_rec_header_t *head)
{
xfs_debug(mp, "%s: SB : uuid = %pU, fmt = %d",
__func__, &mp->m_sb.sb_uuid, XLOG_FMT);
xfs_debug(mp, " log : uuid = %pU, fmt = %d",
&head->h_fs_uuid, be32_to_cpu(head->h_fmt));
}
#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(head->h_magicno == cpu_to_be32(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(head->h_fmt != cpu_to_be32(XLOG_FMT))) {
xfs_warn(mp,
"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 -EFSCORRUPTED;
} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
xfs_warn(mp,
"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 -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(head->h_magicno == cpu_to_be32(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.
*/
xfs_warn(mp, "nil uuid in log - IRIX style log");
} else if (unlikely(!uuid_equal(&mp->m_sb.sb_uuid, &head->h_fs_uuid))) {
xfs_warn(mp, "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 -EFSCORRUPTED;
}
return 0;
}
STATIC void
xlog_recover_iodone(
struct xfs_buf *bp)
{
if (bp->b_error) {
/*
* We're not going to bother about retrying
* this during recovery. One strike!
*/
if (!XFS_FORCED_SHUTDOWN(bp->b_target->bt_mount)) {
xfs_buf_ioerror_alert(bp, __func__);
xfs_force_shutdown(bp->b_target->bt_mount,
SHUTDOWN_META_IO_ERROR);
}
}
/*
* On v5 supers, a bli could be attached to update the metadata LSN.
* Clean it up.
*/
if (bp->b_fspriv)
xfs_buf_item_relse(bp);
ASSERT(bp->b_fspriv == NULL);
bp->b_iodone = NULL;
xfs_buf_ioend(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.
*/
STATIC int
xlog_find_cycle_start(
struct xlog *log,
struct xfs_buf *bp,
xfs_daddr_t first_blk,
xfs_daddr_t *last_blk,
uint cycle)
{
char *offset;
xfs_daddr_t mid_blk;
xfs_daddr_t end_blk;
uint mid_cycle;
int error;
end_blk = *last_blk;
mid_blk = BLK_AVG(first_blk, end_blk);
while (mid_blk != first_blk && mid_blk != end_blk) {
error = xlog_bread(log, mid_blk, 1, bp, &offset);
if (error)
return error;
mid_cycle = xlog_get_cycle(offset);
if (mid_cycle == cycle)
end_blk = mid_blk; /* last_half_cycle == mid_cycle */
else
first_blk = mid_blk; /* first_half_cycle == mid_cycle */
mid_blk = BLK_AVG(first_blk, end_blk);
}
ASSERT((mid_blk == first_blk && mid_blk+1 == end_blk) ||
(mid_blk == end_blk && mid_blk-1 == first_blk));
*last_blk = end_blk;
return 0;
}
/*
* Check that a range of blocks does not contain stop_on_cycle_no.
* Fill in *new_blk with the block offset where such a block is
* found, or with -1 (an invalid block number) if there is no such
* block in the range. The scan needs to occur from front to back
* and the pointer into the region must be updated since a later
* routine will need to perform another test.
*/
STATIC int
xlog_find_verify_cycle(
struct xlog *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;
char *buf = NULL;
int error = 0;
/*
* Greedily allocate a buffer big enough to handle the full
* range of basic blocks we'll be examining. If that fails,
* try a smaller size. We need to be able to read at least
* a log sector, or we're out of luck.
*/
bufblks = 1 << ffs(nbblks);
while (bufblks > log->l_logBBsize)
bufblks >>= 1;
while (!(bp = xlog_get_bp(log, bufblks))) {
bufblks >>= 1;
if (bufblks < log->l_sectBBsize)
return -ENOMEM;
}
for (i = start_blk; i < start_blk + nbblks; i += bufblks) {
int bcount;
bcount = min(bufblks, (start_blk + nbblks - i));
error = xlog_bread(log, i, bcount, bp, &buf);
if (error)
goto out;
for (j = 0; j < bcount; j++) {
cycle = xlog_get_cycle(buf);
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(
struct xlog *log,
xfs_daddr_t start_blk,
xfs_daddr_t *last_blk,
int extra_bblks)
{
xfs_daddr_t i;
xfs_buf_t *bp;
char *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 {
error = xlog_bread(log, start_blk, num_blks, bp, &offset);
if (error)
goto out;
offset += ((num_blks - 1) << BBSHIFT);
}
for (i = (*last_blk) - 1; i >= 0; i--) {
if (i < start_blk) {
/* valid log record not found */
xfs_warn(log->l_mp,
"Log inconsistent (didn't find previous header)");
ASSERT(0);
error = -EIO;
goto out;
}
if (smallmem) {
error = xlog_bread(log, i, 1, bp, &offset);
if (error)
goto out;
}
head = (xlog_rec_header_t *)offset;
if (head->h_magicno == cpu_to_be32(XLOG_HEADER_MAGIC_NUM))
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) {
error = 1;
goto out;
}
/*
* We have the final block of the good log (the first block
* of the log record _before_ the head. So we check the uuid.
*/
if ((error = xlog_header_check_mount(log->l_mp, head)))
goto out;
/*
* We may have found a log record header before we expected one.
* last_blk will be the 1st block # with a given cycle #. We may end
* up reading an entire log record. In this case, we don't want to
* reset last_blk. Only when last_blk points in the middle of a log
* record do we update last_blk.
*/
if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) {
uint h_size = be32_to_cpu(head->h_size);
xhdrs = h_size / XLOG_HEADER_CYCLE_SIZE;
if (h_size % XLOG_HEADER_CYCLE_SIZE)
xhdrs++;
} else {
xhdrs = 1;
}
if (*last_blk - i + extra_bblks !=
BTOBB(be32_to_cpu(head->h_len)) + xhdrs)
*last_blk = i;
out:
xlog_put_bp(bp);
return error;
}
/*
* Head is defined to be the point of the log where the next log write
* could go. This means that incomplete LR writes at the end are
* eliminated when calculating the head. We aren't guaranteed that previous
* LR have complete transactions. We only know that a cycle number of
* current cycle number -1 won't be present in the log if we start writing
* from our current block number.
*
* last_blk contains the block number of the first block with a given
* cycle number.
*
* Return: zero if normal, non-zero if error.
*/
STATIC int
xlog_find_head(
struct xlog *log,
xfs_daddr_t *return_head_blk)
{
xfs_buf_t *bp;
char *offset;
xfs_daddr_t new_blk, first_blk, start_blk, last_blk, head_blk;
int num_scan_bblks;
uint first_half_cycle, last_half_cycle;
uint stop_on_cycle;
int error, log_bbnum = log->l_logBBsize;
/* Is the end of the log device zeroed? */
error = xlog_find_zeroed(log, &first_blk);
if (error < 0) {
xfs_warn(log->l_mp, "empty log check failed");
return error;
}
if (error == 1) {
*return_head_blk = first_blk;
/* Is the whole lot zeroed? */
if (!first_blk) {
/* Linux XFS shouldn't generate totally zeroed logs -
* mkfs etc write a dummy unmount record to a fresh
* log so we can store the uuid in there
*/
xfs_warn(log->l_mp, "totally zeroed log");
}
return 0;
}
first_blk = 0; /* get cycle # of 1st block */
bp = xlog_get_bp(log, 1);
if (!bp)
return -ENOMEM;
error = xlog_bread(log, 0, 1, bp, &offset);
if (error)
goto bp_err;
first_half_cycle = xlog_get_cycle(offset);
last_blk = head_blk = log_bbnum - 1; /* get cycle # of last block */
error = xlog_bread(log, last_blk, 1, bp, &offset);
if (error)
goto bp_err;
last_half_cycle = xlog_get_cycle(offset);
ASSERT(last_half_cycle != 0);
/*
* If the 1st half cycle number is equal to the last half cycle number,
* then the entire log is stamped with the same cycle number. In this
* case, head_blk can't be set to zero (which makes sense). The below
* math doesn't work out properly with head_blk equal to zero. Instead,
* we set it to log_bbnum which is an invalid block number, but this
* value makes the math correct. If head_blk doesn't changed through
* all the tests below, *head_blk is set to zero at the very end rather
* than log_bbnum. In a sense, log_bbnum and zero are the same block
* in a circular file.
*/
if (first_half_cycle == last_half_cycle) {
/*
* In this case we believe that the entire log should have
* cycle number last_half_cycle. We need to scan backwards
* from the end verifying that there are no holes still
* containing last_half_cycle - 1. If we find such a hole,
* then the start of that hole will be the new head. The
* simple case looks like
* x | x ... | x - 1 | x
* Another case that fits this picture would be
* x | x + 1 | x ... | x
* In this case the head really is somewhere at the end of the
* log, as one of the latest writes at the beginning was
* incomplete.
* One more case is
* x | x + 1 | x ... | x - 1 | x
* This is really the combination of the above two cases, and
* the head has to end up at the start of the x-1 hole at the
* end of the log.
*
* In the 256k log case, we will read from the beginning to the
* end of the log and search for cycle numbers equal to x-1.
* We don't worry about the x+1 blocks that we encounter,
* because we know that they cannot be the head since the log
* started with x.
*/
head_blk = log_bbnum;
stop_on_cycle = last_half_cycle - 1;
} else {
/*
* In this case we want to find the first block with cycle
* number matching last_half_cycle. We expect the log to be
* some variation on
* x + 1 ... | x ... | x
* The first block with cycle number x (last_half_cycle) will
* be where the new head belongs. First we do a binary search
* for the first occurrence of last_half_cycle. The binary
* search may not be totally accurate, so then we scan back
* from there looking for occurrences of last_half_cycle before
* us. If that backwards scan wraps around the beginning of
* the log, then we look for occurrences of last_half_cycle - 1
* at the end of the log. The cases we're looking for look
* like
* v binary search stopped here
* x + 1 ... | x | x + 1 | x ... | x
* ^ but we want to locate this spot
* or
* <---------> less than scan distance
* x + 1 ... | x ... | x - 1 | x
* ^ we want to locate this spot
*/
stop_on_cycle = last_half_cycle;
if ((error = xlog_find_cycle_start(log, bp, first_blk,
&head_blk, last_half_cycle)))
goto bp_err;
}
/*
* Now validate the answer. Scan back some number of maximum possible
* blocks and make sure each one has the expected cycle number. The
* maximum is determined by the total possible amount of buffering
* in the in-core log. The following number can be made tighter if
* we actually look at the block size of the filesystem.
*/
num_scan_bblks = XLOG_TOTAL_REC_SHIFT(log);
if (head_blk >= num_scan_bblks) {
/*
* We are guaranteed that the entire check can be performed
* in one buffer.
*/
start_blk = head_blk - num_scan_bblks;
if ((error = xlog_find_verify_cycle(log,
start_blk, num_scan_bblks,
stop_on_cycle, &new_blk)))
goto bp_err;
if (new_blk != -1)
head_blk = new_blk;
} else { /* need to read 2 parts of log */
/*
* We are going to scan backwards in the log in two parts.
* First we scan the physical end of the log. In this part
* of the log, we are looking for blocks with cycle number
* last_half_cycle - 1.
* If we find one, then we know that the log starts there, as
* we've found a hole that didn't get written in going around
* the end of the physical log. The simple case for this is
* x + 1 ... | x ... | x - 1 | x
* <---------> less than scan distance
* If all of the blocks at the end of the log have cycle number
* last_half_cycle, then we check the blocks at the start of
* the log looking for occurrences of last_half_cycle. If we
* find one, then our current estimate for the location of the
* first occurrence of last_half_cycle is wrong and we move
* back to the hole we've found. This case looks like
* x + 1 ... | x | x + 1 | x ...
* ^ binary search stopped here
* Another case we need to handle that only occurs in 256k
* logs is
* x + 1 ... | x ... | x+1 | x ...
* ^ binary search stops here
* In a 256k log, the scan at the end of the log will see the
* x + 1 blocks. We need to skip past those since that is
* certainly not the head of the log. By searching for
* last_half_cycle-1 we accomplish that.
*/
ASSERT(head_blk <= INT_MAX &&
(xfs_daddr_t) num_scan_bblks >= head_blk);
start_blk = log_bbnum - (num_scan_bblks - head_blk);
if ((error = xlog_find_verify_cycle(log, start_blk,
num_scan_bblks - (int)head_blk,
(stop_on_cycle - 1), &new_blk)))
goto bp_err;
if (new_blk != -1) {
head_blk = new_blk;
goto validate_head;
}
/*
* Scan beginning of log now. The last part of the physical
* log is good. This scan needs to verify that it doesn't find
* the last_half_cycle.
*/
start_blk = 0;
ASSERT(head_blk <= INT_MAX);
if ((error = xlog_find_verify_cycle(log,
start_blk, (int)head_blk,
stop_on_cycle, &new_blk)))
goto bp_err;
if (new_blk != -1)
head_blk = new_blk;
}
validate_head:
/*
* Now we need to make sure head_blk is not pointing to a block in
* the middle of a log record.
*/
num_scan_bblks = XLOG_REC_SHIFT(log);
if (head_blk >= num_scan_bblks) {
start_blk = head_blk - num_scan_bblks; /* don't read head_blk */
/* start ptr at last block ptr before head_blk */
error = xlog_find_verify_log_record(log, start_blk, &head_blk, 0);
if (error == 1)
error = -EIO;
if (error)
goto bp_err;
} else {
start_blk = 0;
ASSERT(head_blk <= INT_MAX);
error = xlog_find_verify_log_record(log, start_blk, &head_blk, 0);
if (error < 0)
goto bp_err;
if (error == 1) {
/* We hit the beginning of the log during our search */
start_blk = log_bbnum - (num_scan_bblks - head_blk);
new_blk = log_bbnum;
ASSERT(start_blk <= INT_MAX &&
(xfs_daddr_t) log_bbnum-start_blk >= 0);
ASSERT(head_blk <= INT_MAX);
error = xlog_find_verify_log_record(log, start_blk,
&new_blk, (int)head_blk);
if (error == 1)
error = -EIO;
if (error)
goto bp_err;
if (new_blk != log_bbnum)
head_blk = new_blk;
} else if (error)
goto bp_err;
}
xlog_put_bp(bp);
if (head_blk == log_bbnum)
*return_head_blk = 0;
else
*return_head_blk = head_blk;
/*
* When returning here, we have a good block number. Bad block
* means that during a previous crash, we didn't have a clean break
* from cycle number N to cycle number N-1. In this case, we need
* to find the first block with cycle number N-1.
*/
return 0;
bp_err:
xlog_put_bp(bp);
if (error)
xfs_warn(log->l_mp, "failed to find log head");
return error;
}
/*
* Seek backwards in the log for log record headers.
*
* Given a starting log block, walk backwards until we find the provided number
* of records or hit the provided tail block. The return value is the number of
* records encountered or a negative error code. The log block and buffer
* pointer of the last record seen are returned in rblk and rhead respectively.
*/
STATIC int
xlog_rseek_logrec_hdr(
struct xlog *log,
xfs_daddr_t head_blk,
xfs_daddr_t tail_blk,
int count,
struct xfs_buf *bp,
xfs_daddr_t *rblk,
struct xlog_rec_header **rhead,
bool *wrapped)
{
int i;
int error;
int found = 0;
char *offset = NULL;
xfs_daddr_t end_blk;
*wrapped = false;
/*
* Walk backwards from the head block until we hit the tail or the first
* block in the log.
*/
end_blk = head_blk > tail_blk ? tail_blk : 0;
for (i = (int) head_blk - 1; i >= end_blk; i--) {
error = xlog_bread(log, i, 1, bp, &offset);
if (error)
goto out_error;
if (*(__be32 *) offset == cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
*rblk = i;
*rhead = (struct xlog_rec_header *) offset;
if (++found == count)
break;
}
}
/*
* If we haven't hit the tail block or the log record header count,
* start looking again from the end of the physical log. Note that
* callers can pass head == tail if the tail is not yet known.
*/
if (tail_blk >= head_blk && found != count) {
for (i = log->l_logBBsize - 1; i >= (int) tail_blk; i--) {
error = xlog_bread(log, i, 1, bp, &offset);
if (error)
goto out_error;
if (*(__be32 *)offset ==
cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
*wrapped = true;
*rblk = i;
*rhead = (struct xlog_rec_header *) offset;
if (++found == count)
break;
}
}
}
return found;
out_error:
return error;
}
/*
* Seek forward in the log for log record headers.
*
* Given head and tail blocks, walk forward from the tail block until we find
* the provided number of records or hit the head block. The return value is the
* number of records encountered or a negative error code. The log block and
* buffer pointer of the last record seen are returned in rblk and rhead
* respectively.
*/
STATIC int
xlog_seek_logrec_hdr(
struct xlog *log,
xfs_daddr_t head_blk,
xfs_daddr_t tail_blk,
int count,
struct xfs_buf *bp,
xfs_daddr_t *rblk,
struct xlog_rec_header **rhead,
bool *wrapped)
{
int i;
int error;
int found = 0;
char *offset = NULL;
xfs_daddr_t end_blk;
*wrapped = false;
/*
* Walk forward from the tail block until we hit the head or the last
* block in the log.
*/
end_blk = head_blk > tail_blk ? head_blk : log->l_logBBsize - 1;
for (i = (int) tail_blk; i <= end_blk; i++) {
error = xlog_bread(log, i, 1, bp, &offset);
if (error)
goto out_error;
if (*(__be32 *) offset == cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
*rblk = i;
*rhead = (struct xlog_rec_header *) offset;
if (++found == count)
break;
}
}
/*
* If we haven't hit the head block or the log record header count,
* start looking again from the start of the physical log.
*/
if (tail_blk > head_blk && found != count) {
for (i = 0; i < (int) head_blk; i++) {
error = xlog_bread(log, i, 1, bp, &offset);
if (error)
goto out_error;
if (*(__be32 *)offset ==
cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) {
*wrapped = true;
*rblk = i;
*rhead = (struct xlog_rec_header *) offset;
if (++found == count)
break;
}
}
}
return found;
out_error:
return error;
}
/*
* Check the log tail for torn writes. This is required when torn writes are
* detected at the head and the head had to be walked back to a previous record.
* The tail of the previous record must now be verified to ensure the torn
* writes didn't corrupt the previous tail.
*
* Return an error if CRC verification fails as recovery cannot proceed.
*/
STATIC int
xlog_verify_tail(
struct xlog *log,
xfs_daddr_t head_blk,
xfs_daddr_t tail_blk)
{
struct xlog_rec_header *thead;
struct xfs_buf *bp;
xfs_daddr_t first_bad;
int count;
int error = 0;
bool wrapped;
xfs_daddr_t tmp_head;
bp = xlog_get_bp(log, 1);
if (!bp)
return -ENOMEM;
/*
* Seek XLOG_MAX_ICLOGS + 1 records past the current tail record to get
* a temporary head block that points after the last possible
* concurrently written record of the tail.
*/
count = xlog_seek_logrec_hdr(log, head_blk, tail_blk,
XLOG_MAX_ICLOGS + 1, bp, &tmp_head, &thead,
&wrapped);
if (count < 0) {
error = count;
goto out;
}
/*
* If the call above didn't find XLOG_MAX_ICLOGS + 1 records, we ran
* into the actual log head. tmp_head points to the start of the record
* so update it to the actual head block.
*/
if (count < XLOG_MAX_ICLOGS + 1)
tmp_head = head_blk;
/*
* We now have a tail and temporary head block that covers at least
* XLOG_MAX_ICLOGS records from the tail. We need to verify that these
* records were completely written. Run a CRC verification pass from
* tail to head and return the result.
*/
error = xlog_do_recovery_pass(log, tmp_head, tail_blk,
XLOG_RECOVER_CRCPASS, &first_bad);
out:
xlog_put_bp(bp);
return error;
}
/*
* Detect and trim torn writes from the head of the log.
*
* Storage without sector atomicity guarantees can result in torn writes in the
* log in the event of a crash. Our only means to detect this scenario is via
* CRC verification. While we can't always be certain that CRC verification
* failure is due to a torn write vs. an unrelated corruption, we do know that
* only a certain number (XLOG_MAX_ICLOGS) of log records can be written out at
* one time. Therefore, CRC verify up to XLOG_MAX_ICLOGS records at the head of
* the log and treat failures in this range as torn writes as a matter of
* policy. In the event of CRC failure, the head is walked back to the last good
* record in the log and the tail is updated from that record and verified.
*/
STATIC int
xlog_verify_head(
struct xlog *log,
xfs_daddr_t *head_blk, /* in/out: unverified head */
xfs_daddr_t *tail_blk, /* out: tail block */
struct xfs_buf *bp,
xfs_daddr_t *rhead_blk, /* start blk of last record */
struct xlog_rec_header **rhead, /* ptr to last record */
bool *wrapped) /* last rec. wraps phys. log */
{
struct xlog_rec_header *tmp_rhead;
struct xfs_buf *tmp_bp;
xfs_daddr_t first_bad;
xfs_daddr_t tmp_rhead_blk;
int found;
int error;
bool tmp_wrapped;
/*
* Check the head of the log for torn writes. Search backwards from the
* head until we hit the tail or the maximum number of log record I/Os
* that could have been in flight at one time. Use a temporary buffer so
* we don't trash the rhead/bp pointers from the caller.
*/
tmp_bp = xlog_get_bp(log, 1);
if (!tmp_bp)
return -ENOMEM;
error = xlog_rseek_logrec_hdr(log, *head_blk, *tail_blk,
XLOG_MAX_ICLOGS, tmp_bp, &tmp_rhead_blk,
&tmp_rhead, &tmp_wrapped);
xlog_put_bp(tmp_bp);
if (error < 0)
return error;
/*
* Now run a CRC verification pass over the records starting at the
* block found above to the current head. If a CRC failure occurs, the
* log block of the first bad record is saved in first_bad.
*/
error = xlog_do_recovery_pass(log, *head_blk, tmp_rhead_blk,
XLOG_RECOVER_CRCPASS, &first_bad);
if (error == -EFSBADCRC) {
/*
* We've hit a potential torn write. Reset the error and warn
* about it.
*/
error = 0;
xfs_warn(log->l_mp,
"Torn write (CRC failure) detected at log block 0x%llx. Truncating head block from 0x%llx.",
first_bad, *head_blk);
/*
* Get the header block and buffer pointer for the last good
* record before the bad record.
*
* Note that xlog_find_tail() clears the blocks at the new head
* (i.e., the records with invalid CRC) if the cycle number
* matches the the current cycle.
*/
found = xlog_rseek_logrec_hdr(log, first_bad, *tail_blk, 1, bp,
rhead_blk, rhead, wrapped);
if (found < 0)
return found;
if (found == 0) /* XXX: right thing to do here? */
return -EIO;
/*
* Reset the head block to the starting block of the first bad
* log record and set the tail block based on the last good
* record.
*
* Bail out if the updated head/tail match as this indicates
* possible corruption outside of the acceptable
* (XLOG_MAX_ICLOGS) range. This is a job for xfs_repair...
*/
*head_blk = first_bad;
*tail_blk = BLOCK_LSN(be64_to_cpu((*rhead)->h_tail_lsn));
if (*head_blk == *tail_blk) {
ASSERT(0);
return 0;
}
/*
* Now verify the tail based on the updated head. This is
* required because the torn writes trimmed from the head could
* have been written over the tail of a previous record. Return
* any errors since recovery cannot proceed if the tail is
* corrupt.
*
* XXX: This leaves a gap in truly robust protection from torn
* writes in the log. If the head is behind the tail, the tail
* pushes forward to create some space and then a crash occurs
* causing the writes into the previous record's tail region to
* tear, log recovery isn't able to recover.
*
* How likely is this to occur? If possible, can we do something
* more intelligent here? Is it safe to push the tail forward if
* we can determine that the tail is within the range of the
* torn write (e.g., the kernel can only overwrite the tail if
* it has actually been pushed forward)? Alternatively, could we
* somehow prevent this condition at runtime?
*/
error = xlog_verify_tail(log, *head_blk, *tail_blk);
}
return error;
}
/*
* Check whether the head of the log points to an unmount record. In other
* words, determine whether the log is clean. If so, update the in-core state
* appropriately.
*/
static int
xlog_check_unmount_rec(
struct xlog *log,
xfs_daddr_t *head_blk,
xfs_daddr_t *tail_blk,
struct xlog_rec_header *rhead,
xfs_daddr_t rhead_blk,
struct xfs_buf *bp,
bool *clean)
{
struct xlog_op_header *op_head;
xfs_daddr_t umount_data_blk;
xfs_daddr_t after_umount_blk;
int hblks;
int error;
char *offset;
*clean = false;
/*
* Look for unmount record. If we find it, then we know there was a
* clean unmount. Since 'i' could be the last block in the physical
* log, we convert to a log block before comparing to the head_blk.
*
* Save the current tail lsn to use to pass to xlog_clear_stale_blocks()
* below. We won't want to clear the unmount record if there is one, so
* we pass the lsn of the unmount record rather than the block after it.
*/
if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) {
int h_size = be32_to_cpu(rhead->h_size);
int h_version = be32_to_cpu(rhead->h_version);
if ((h_version & XLOG_VERSION_2) &&
(h_size > XLOG_HEADER_CYCLE_SIZE)) {
hblks = h_size / XLOG_HEADER_CYCLE_SIZE;
if (h_size % XLOG_HEADER_CYCLE_SIZE)
hblks++;
} else {
hblks = 1;
}
} else {
hblks = 1;
}
after_umount_blk = rhead_blk + hblks + BTOBB(be32_to_cpu(rhead->h_len));
after_umount_blk = do_mod(after_umount_blk, log->l_logBBsize);
if (*head_blk == after_umount_blk &&
be32_to_cpu(rhead->h_num_logops) == 1) {
umount_data_blk = rhead_blk + hblks;
umount_data_blk = do_mod(umount_data_blk, log->l_logBBsize);
error = xlog_bread(log, umount_data_blk, 1, bp, &offset);
if (error)
return error;
op_head = (struct xlog_op_header *)offset;
if (op_head->oh_flags & XLOG_UNMOUNT_TRANS) {
/*
* Set tail and last sync so that newly written log
* records will point recovery to after the current
* unmount record.
*/
xlog_assign_atomic_lsn(&log->l_tail_lsn,
log->l_curr_cycle, after_umount_blk);
xlog_assign_atomic_lsn(&log->l_last_sync_lsn,
log->l_curr_cycle, after_umount_blk);
*tail_blk = after_umount_blk;
*clean = true;
}
}
return 0;
}
static void
xlog_set_state(
struct xlog *log,
xfs_daddr_t head_blk,
struct xlog_rec_header *rhead,
xfs_daddr_t rhead_blk,
bool bump_cycle)
{
/*
* Reset log values according to the state of the log when we
* crashed. In the case where head_blk == 0, we bump curr_cycle
* one because the next write starts a new cycle rather than
* continuing the cycle of the last good log record. At this
* point we have guaranteed that all partial log records have been
* accounted for. Therefore, we know that the last good log record
* written was complete and ended exactly on the end boundary
* of the physical log.
*/
log->l_prev_block = rhead_blk;
log->l_curr_block = (int)head_blk;
log->l_curr_cycle = be32_to_cpu(rhead->h_cycle);
if (bump_cycle)
log->l_curr_cycle++;
atomic64_set(&log->l_tail_lsn, be64_to_cpu(rhead->h_tail_lsn));
atomic64_set(&log->l_last_sync_lsn, be64_to_cpu(rhead->h_lsn));
xlog_assign_grant_head(&log->l_reserve_head.grant, log->l_curr_cycle,
BBTOB(log->l_curr_block));
xlog_assign_grant_head(&log->l_write_head.grant, log->l_curr_cycle,
BBTOB(log->l_curr_block));
}
/*
* Find the sync block number or the tail of the log.
*
* This will be the block number of the last record to have its
* associated buffers synced to disk. Every log record header has
* a sync lsn embedded in it. LSNs hold block numbers, so it is easy
* to get a sync block number. The only concern is to figure out which
* log record header to believe.
*
* The following algorithm uses the log record header with the largest
* lsn. The entire log record does not need to be valid. We only care
* that the header is valid.
*
* We could speed up search by using current head_blk buffer, but it is not
* available.
*/
STATIC int
xlog_find_tail(
struct xlog *log,
xfs_daddr_t *head_blk,
xfs_daddr_t *tail_blk)
{
xlog_rec_header_t *rhead;
char *offset = NULL;
xfs_buf_t *bp;
int error;
xfs_daddr_t rhead_blk;
xfs_lsn_t tail_lsn;
bool wrapped = false;
bool clean = false;
/*
* Find previous log record
*/
if ((error = xlog_find_head(log, head_blk)))
return error;
ASSERT(*head_blk < INT_MAX);
bp = xlog_get_bp(log, 1);
if (!bp)
return -ENOMEM;
if (*head_blk == 0) { /* special case */
error = xlog_bread(log, 0, 1, bp, &offset);
if (error)
goto done;
if (xlog_get_cycle(offset) == 0) {
*tail_blk = 0;
/* leave all other log inited values alone */
goto done;
}
}
/*
* Search backwards through the log looking for the log record header
* block. This wraps all the way back around to the head so something is
* seriously wrong if we can't find it.
*/
error = xlog_rseek_logrec_hdr(log, *head_blk, *head_blk, 1, bp,
&rhead_blk, &rhead, &wrapped);
if (error < 0)
return error;
if (!error) {
xfs_warn(log->l_mp, "%s: couldn't find sync record", __func__);
return -EIO;
}
*tail_blk = BLOCK_LSN(be64_to_cpu(rhead->h_tail_lsn));
/*
* Set the log state based on the current head record.
*/
xlog_set_state(log, *head_blk, rhead, rhead_blk, wrapped);
tail_lsn = atomic64_read(&log->l_tail_lsn);
/*
* Look for an unmount record at the head of the log. This sets the log
* state to determine whether recovery is necessary.
*/
error = xlog_check_unmount_rec(log, head_blk, tail_blk, rhead,
rhead_blk, bp, &clean);
if (error)
goto done;
/*
* Verify the log head if the log is not clean (e.g., we have anything
* but an unmount record at the head). This uses CRC verification to
* detect and trim torn writes. If discovered, CRC failures are
* considered torn writes and the log head is trimmed accordingly.
*
* Note that we can only run CRC verification when the log is dirty
* because there's no guarantee that the log data behind an unmount
* record is compatible with the current architecture.
*/
if (!clean) {
xfs_daddr_t orig_head = *head_blk;
error = xlog_verify_head(log, head_blk, tail_blk, bp,
&rhead_blk, &rhead, &wrapped);
if (error)
goto done;
/* update in-core state again if the head changed */
if (*head_blk != orig_head) {
xlog_set_state(log, *head_blk, rhead, rhead_blk,
wrapped);
tail_lsn = atomic64_read(&log->l_tail_lsn);
error = xlog_check_unmount_rec(log, head_blk, tail_blk,
rhead, rhead_blk, bp,
&clean);
if (error)
goto done;
}
}
/*
* Note that the unmount was clean. If the unmount was not clean, we
* need to know this to rebuild the superblock counters from the perag
* headers if we have a filesystem using non-persistent counters.
*/
if (clean)
log->l_mp->m_flags |= XFS_MOUNT_WAS_CLEAN;
/*
* Make sure that there are no blocks in front of the head
* with the same cycle number as the head. This can happen
* because we allow multiple outstanding log writes concurrently,
* and the later writes might make it out before earlier ones.
*
* We use the lsn from before modifying it so that we'll never
* overwrite the unmount record after a clean unmount.
*
* Do this only if we are going to recover the filesystem
*
* NOTE: This used to say "if (!readonly)"
* However on Linux, we can & do recover a read-only filesystem.
* We only skip recovery if NORECOVERY is specified on mount,
* in which case we would not be here.
*
* But... if the -device- itself is readonly, just skip this.
* We can't recover this device anyway, so it won't matter.
*/
if (!xfs_readonly_buftarg(log->l_mp->m_logdev_targp))
error = xlog_clear_stale_blocks(log, tail_lsn);
done:
xlog_put_bp(bp);
if (error)
xfs_warn(log->l_mp, "failed to locate log tail");
return error;
}
/*
* Is the log zeroed at all?
*
* The last binary search should be changed to perform an X block read
* once X becomes small enough. You can then search linearly through
* the X blocks. This will cut down on the number of reads we need to do.
*
* If the log is partially zeroed, this routine will pass back the blkno
* of the first block with cycle number 0. It won't have a complete LR
* preceding it.
*
* Return:
* 0 => the log is completely written to
* 1 => use *blk_no as the first block of the log
* <0 => error has occurred
*/
STATIC int
xlog_find_zeroed(
struct xlog *log,
xfs_daddr_t *blk_no)
{
xfs_buf_t *bp;
char *offset;
uint first_cycle, last_cycle;
xfs_daddr_t new_blk, last_blk, start_blk;
xfs_daddr_t num_scan_bblks;
int error, log_bbnum = log->l_logBBsize;
*blk_no = 0;
/* check totally zeroed log */
bp = xlog_get_bp(log, 1);
if (!bp)
return -ENOMEM;
error = xlog_bread(log, 0, 1, bp, &offset);
if (error)
goto bp_err;
first_cycle = xlog_get_cycle(offset);
if (first_cycle == 0) { /* completely zeroed log */
*blk_no = 0;
xlog_put_bp(bp);
return 1;
}
/* check partially zeroed log */
error = xlog_bread(log, log_bbnum-1, 1, bp, &offset);
if (error)
goto bp_err;
last_cycle = xlog_get_cycle(offset);
if (last_cycle != 0) { /* log completely written to */
xlog_put_bp(bp);
return 0;
} else if (first_cycle != 1) {
/*
* If the cycle of the last block is zero, the cycle of
* the first block must be 1. If it's not, maybe we're
* not looking at a log... Bail out.
*/
xfs_warn(log->l_mp,
"Log inconsistent or not a log (last==0, first!=1)");
error = -EINVAL;
goto bp_err;
}
/* we have a partially zeroed log */
last_blk = log_bbnum-1;
if ((error = xlog_find_cycle_start(log, bp, 0, &last_blk, 0)))
goto bp_err;
/*
* Validate the answer. Because there is no way to guarantee that
* the entire log is made up of log records which are the same size,
* we scan over the defined maximum blocks. At this point, the maximum
* is not chosen to mean anything special. XXXmiken
*/
num_scan_bblks = XLOG_TOTAL_REC_SHIFT(log);
ASSERT(num_scan_bblks <= INT_MAX);
if (last_blk < num_scan_bblks)
num_scan_bblks = last_blk;
start_blk = last_blk - num_scan_bblks;
/*
* We search for any instances of cycle number 0 that occur before
* our current estimate of the head. What we're trying to detect is
* 1 ... | 0 | 1 | 0...
* ^ binary search ends here
*/
if ((error = xlog_find_verify_cycle(log, start_blk,
(int)num_scan_bblks, 0, &new_blk)))
goto bp_err;
if (new_blk != -1)
last_blk = new_blk;
/*
* Potentially backup over partial log record write. We don't need
* to search the end of the log because we know it is zero.
*/
error = xlog_find_verify_log_record(log, start_blk, &last_blk, 0);
if (error == 1)
error = -EIO;
if (error)
goto bp_err;
*blk_no = last_blk;
bp_err:
xlog_put_bp(bp);
if (error)
return error;
return 1;
}
/*
* These are simple subroutines used by xlog_clear_stale_blocks() below
* to initialize a buffer full of empty log record headers and write
* them into the log.
*/
STATIC void
xlog_add_record(
struct xlog *log,
char *buf,
int cycle,
int block,
int tail_cycle,
int tail_block)
{
xlog_rec_header_t *recp = (xlog_rec_header_t *)buf;
memset(buf, 0, BBSIZE);
recp->h_magicno = cpu_to_be32(XLOG_HEADER_MAGIC_NUM);
recp->h_cycle = cpu_to_be32(cycle);
recp->h_version = cpu_to_be32(
xfs_sb_version_haslogv2(&log->l_mp->m_sb) ? 2 : 1);
recp->h_lsn = cpu_to_be64(xlog_assign_lsn(cycle, block));
recp->h_tail_lsn = cpu_to_be64(xlog_assign_lsn(tail_cycle, tail_block));
recp->h_fmt = cpu_to_be32(XLOG_FMT);
memcpy(&recp->h_fs_uuid, &log->l_mp->m_sb.sb_uuid, sizeof(uuid_t));
}
STATIC int
xlog_write_log_records(
struct xlog *log,
int cycle,
int start_block,
int blocks,
int tail_cycle,
int tail_block)
{
char *offset;
xfs_buf_t *bp;
int balign, ealign;
int sectbb = log->l_sectBBsize;
int end_block = start_block + blocks;
int bufblks;
int error = 0;
int i, j = 0;
/*
* Greedily allocate a buffer big enough to handle the full
* range of basic blocks to be written. If that fails, try
* a smaller size. We need to be able to write at least a
* log sector, or we're out of luck.
*/
bufblks = 1 << ffs(blocks);
while (bufblks > log->l_logBBsize)
bufblks >>= 1;
while (!(bp = xlog_get_bp(log, bufblks))) {
bufblks >>= 1;
if (bufblks < sectbb)
return -ENOMEM;
}
/* We may need to do a read at the start to fill in part of
* the buffer in the starting sector not covered by the first
* write below.
*/
balign = round_down(start_block, sectbb);
if (balign != start_block) {
error = xlog_bread_noalign(log, start_block, 1, bp);
if (error)
goto out_put_bp;
j = start_block - balign;
}
for (i = start_block; i < end_block; i += bufblks) {
int bcount, endcount;
bcount = min(bufblks, end_block - start_block);
endcount = bcount - j;
/* We may need to do a read at the end to fill in part of
* the buffer in the final sector not covered by the write.
* If this is the same sector as the above read, skip it.
*/
ealign = round_down(end_block, sectbb);
if (j == 0 && (start_block + endcount > ealign)) {
offset = bp->b_addr + BBTOB(ealign - start_block);
error = xlog_bread_offset(log, ealign, sectbb,
bp, offset);
if (error)
break;
}
offset = xlog_align(log, start_block, endcount, bp);
for (; j < endcount; j++) {
xlog_add_record(log, offset, cycle, i+j,
tail_cycle, tail_block);
offset += BBSIZE;
}
error = xlog_bwrite(log, start_block, endcount, bp);
if (error)
break;
start_block += endcount;
j = 0;
}
out_put_bp:
xlog_put_bp(bp);
return error;
}
/*
* This routine is called to blow away any incomplete log writes out
* in front of the log head. We do this so that we won't become confused
* if we come up, write only a little bit more, and then crash again.
* If we leave the partial log records out there, this situation could
* cause us to think those partial writes are valid blocks since they
* have the current cycle number. We get rid of them by overwriting them
* with empty log records with the old cycle number rather than the
* current one.
*
* The tail lsn is passed in rather than taken from
* the log so that we will not write over the unmount record after a
* clean unmount in a 512 block log. Doing so would leave the log without
* any valid log records in it until a new one was written. If we crashed
* during that time we would not be able to recover.
*/
STATIC int
xlog_clear_stale_blocks(
struct xlog *log,
xfs_lsn_t tail_lsn)
{
int tail_cycle, head_cycle;
int tail_block, head_block;
int tail_distance, max_distance;
int distance;
int error;
tail_cycle = CYCLE_LSN(tail_lsn);
tail_block = BLOCK_LSN(tail_lsn);
head_cycle = log->l_curr_cycle;
head_block = log->l_curr_block;
/*
* Figure out the distance between the new head of the log
* and the tail. We want to write over any blocks beyond the
* head that we may have written just before the crash, but
* we don't want to overwrite the tail of the log.
*/
if (head_cycle == tail_cycle) {
/*
* The tail is behind the head in the physical log,
* so the distance from the head to the tail is the
* distance from the head to the end of the log plus
* the distance from the beginning of the log to the
* tail.
*/
if (unlikely(head_block < tail_block || head_block >= log->l_logBBsize)) {
XFS_ERROR_REPORT("xlog_clear_stale_blocks(1)",
XFS_ERRLEVEL_LOW, log->l_mp);
return -EFSCORRUPTED;
}
tail_distance = tail_block + (log->l_logBBsize - head_block);
} else {
/*
* The head is behind the tail in the physical log,
* so the distance from the head to the tail is just
* the tail block minus the head block.
*/
if (unlikely(head_block >= tail_block || head_cycle != (tail_cycle + 1))){
XFS_ERROR_REPORT("xlog_clear_stale_blocks(2)",
XFS_ERRLEVEL_LOW, log->l_mp);
return -EFSCORRUPTED;
}
tail_distance = tail_block - head_block;
}
/*
* If the head is right up against the tail, we can't clear
* anything.
*/
if (tail_distance <= 0) {
ASSERT(tail_distance == 0);
return 0;
}
max_distance = XLOG_TOTAL_REC_SHIFT(log);
/*
* Take the smaller of the maximum amount of outstanding I/O
* we could have and the distance to the tail to clear out.
* We take the smaller so that we don't overwrite the tail and
* we don't waste all day writing from the head to the tail
* for no reason.
*/
max_distance = MIN(max_distance, tail_distance);
if ((head_block + max_distance) <= log->l_logBBsize) {
/*
* We can stomp all the blocks we need to without
* wrapping around the end of the log. Just do it
* in a single write. Use the cycle number of the
* current cycle minus one so that the log will look like:
* n ... | n - 1 ...
*/
error = xlog_write_log_records(log, (head_cycle - 1),
head_block, max_distance, tail_cycle,
tail_block);
if (error)
return error;
} else {
/*
* We need to wrap around the end of the physical log in
* order to clear all the blocks. Do it in two separate
* I/Os. The first write should be from the head to the
* end of the physical log, and it should use the current
* cycle number minus one just like above.
*/
distance = log->l_logBBsize - head_block;
error = xlog_write_log_records(log, (head_cycle - 1),
head_block, distance, tail_cycle,
tail_block);
if (error)
return error;
/*
* Now write the blocks at the start of the physical log.
* This writes the remainder of the blocks we want to clear.
* It uses the current cycle number since we're now on the
* same cycle as the head so that we get:
* n ... n ... | n - 1 ...
* ^^^^^ blocks we're writing
*/
distance = max_distance - (log->l_logBBsize - head_block);
error = xlog_write_log_records(log, head_cycle, 0, distance,
tail_cycle, tail_block);
if (error)
return error;
}
return 0;
}
/******************************************************************************
*
* Log recover routines
*
******************************************************************************
*/
/*
* Sort the log items in the transaction.
*
* The ordering constraints are defined by the inode allocation and unlink
* behaviour. The rules are:
*
* 1. Every item is only logged once in a given transaction. Hence it
* represents the last logged state of the item. Hence ordering is
* dependent on the order in which operations need to be performed so
* required initial conditions are always met.
*
* 2. Cancelled buffers are recorded in pass 1 in a separate table and
* there's nothing to replay from them so we can simply cull them
* from the transaction. However, we can't do that until after we've
* replayed all the other items because they may be dependent on the
* cancelled buffer and replaying the cancelled buffer can remove it
* form the cancelled buffer table. Hence they have tobe done last.
*
* 3. Inode allocation buffers must be replayed before inode items that
* read the buffer and replay changes into it. For filesystems using the
* ICREATE transactions, this means XFS_LI_ICREATE objects need to get
* treated the same as inode allocation buffers as they create and
* initialise the buffers directly.
*
* 4. Inode unlink buffers must be replayed after inode items are replayed.
* This ensures that inodes are completely flushed to the inode buffer
* in a "free" state before we remove the unlinked inode list pointer.
*
* Hence the ordering needs to be inode allocation buffers first, inode items
* second, inode unlink buffers third and cancelled buffers last.
*
* But there's a problem with that - we can't tell an inode allocation buffer
* apart from a regular buffer, so we can't separate them. We can, however,
* tell an inode unlink buffer from the others, and so we can separate them out
* from all the other buffers and move them to last.
*
* Hence, 4 lists, in order from head to tail:
* - buffer_list for all buffers except cancelled/inode unlink buffers
* - item_list for all non-buffer items
* - inode_buffer_list for inode unlink buffers
* - cancel_list for the cancelled buffers
*
* Note that we add objects to the tail of the lists so that first-to-last
* ordering is preserved within the lists. Adding objects to the head of the
* list means when we traverse from the head we walk them in last-to-first
* order. For cancelled buffers and inode unlink buffers this doesn't matter,
* but for all other items there may be specific ordering that we need to
* preserve.
*/
STATIC int
xlog_recover_reorder_trans(
struct xlog *log,
struct xlog_recover *trans,
int pass)
{
xlog_recover_item_t *item, *n;
int error = 0;
LIST_HEAD(sort_list);
LIST_HEAD(cancel_list);
LIST_HEAD(buffer_list);
LIST_HEAD(inode_buffer_list);
LIST_HEAD(inode_list);
list_splice_init(&trans->r_itemq, &sort_list);
list_for_each_entry_safe(item, n, &sort_list, ri_list) {
xfs_buf_log_format_t *buf_f = item->ri_buf[0].i_addr;
switch (ITEM_TYPE(item)) {
case XFS_LI_ICREATE:
list_move_tail(&item->ri_list, &buffer_list);
break;
case XFS_LI_BUF:
if (buf_f->blf_flags & XFS_BLF_CANCEL) {
trace_xfs_log_recover_item_reorder_head(log,
trans, item, pass);
list_move(&item->ri_list, &cancel_list);
break;
}
if (buf_f->blf_flags & XFS_BLF_INODE_BUF) {
list_move(&item->ri_list, &inode_buffer_list);
break;
}
list_move_tail(&item->ri_list, &buffer_list);
break;
case XFS_LI_INODE:
case XFS_LI_DQUOT:
case XFS_LI_QUOTAOFF:
case XFS_LI_EFD:
case XFS_LI_EFI:
case XFS_LI_RUI:
case XFS_LI_RUD:
case XFS_LI_CUI:
case XFS_LI_CUD:
case XFS_LI_BUI:
case XFS_LI_BUD:
trace_xfs_log_recover_item_reorder_tail(log,
trans, item, pass);
list_move_tail(&item->ri_list, &inode_list);
break;
default:
xfs_warn(log->l_mp,
"%s: unrecognized type of log operation",
__func__);
ASSERT(0);
/*
* return the remaining items back to the transaction
* item list so they can be freed in caller.
*/
if (!list_empty(&sort_list))
list_splice_init(&sort_list, &trans->r_itemq);
error = -EIO;
goto out;
}
}
out:
ASSERT(list_empty(&sort_list));
if (!list_empty(&buffer_list))
list_splice(&buffer_list, &trans->r_itemq);
if (!list_empty(&inode_list))
list_splice_tail(&inode_list, &trans->r_itemq);
if (!list_empty(&inode_buffer_list))
list_splice_tail(&inode_buffer_list, &trans->r_itemq);
if (!list_empty(&cancel_list))
list_splice_tail(&cancel_list, &trans->r_itemq);
return error;
}
/*
* Build up the table of buf cancel records so that we don't replay
* cancelled data in the second pass. For buffer records that are
* not cancel records, there is nothing to do here so we just return.
*
* If we get a cancel record which is already in the table, this indicates
* that the buffer was cancelled multiple times. In order to ensure
* that during pass 2 we keep the record in the table until we reach its
* last occurrence in the log, we keep a reference count in the cancel
* record in the table to tell us how many times we expect to see this
* record during the second pass.
*/
STATIC int
xlog_recover_buffer_pass1(
struct xlog *log,
struct xlog_recover_item *item)
{
xfs_buf_log_format_t *buf_f = item->ri_buf[0].i_addr;
struct list_head *bucket;
struct xfs_buf_cancel *bcp;
/*
* If this isn't a cancel buffer item, then just return.
*/
if (!(buf_f->blf_flags & XFS_BLF_CANCEL)) {
trace_xfs_log_recover_buf_not_cancel(log, buf_f);
return 0;
}
/*
* Insert an xfs_buf_cancel record into the hash table of them.
* If there is already an identical record, bump its reference count.
*/
bucket = XLOG_BUF_CANCEL_BUCKET(log, buf_f->blf_blkno);
list_for_each_entry(bcp, bucket, bc_list) {
if (bcp->bc_blkno == buf_f->blf_blkno &&
bcp->bc_len == buf_f->blf_len) {
bcp->bc_refcount++;
trace_xfs_log_recover_buf_cancel_ref_inc(log, buf_f);
return 0;
}
}
bcp = kmem_alloc(sizeof(struct xfs_buf_cancel), KM_SLEEP);
bcp->bc_blkno = buf_f->blf_blkno;
bcp->bc_len = buf_f->blf_len;
bcp->bc_refcount = 1;
list_add_tail(&bcp->bc_list, bucket);
trace_xfs_log_recover_buf_cancel_add(log, buf_f);
return 0;
}
/*
* Check to see whether the buffer being recovered has a corresponding
* entry in the buffer cancel record table. If it is, return the cancel
* buffer structure to the caller.
*/
STATIC struct xfs_buf_cancel *
xlog_peek_buffer_cancelled(
struct xlog *log,
xfs_daddr_t blkno,
uint len,
unsigned short flags)
{
struct list_head *bucket;
struct xfs_buf_cancel *bcp;
if (!log->l_buf_cancel_table) {
/* empty table means no cancelled buffers in the log */
ASSERT(!(flags & XFS_BLF_CANCEL));
return NULL;
}
bucket = XLOG_BUF_CANCEL_BUCKET(log, blkno);
list_for_each_entry(bcp, bucket, bc_list) {
if (bcp->bc_blkno == blkno && bcp->bc_len == len)
return bcp;
}
/*
* We didn't find a corresponding entry in the table, so return 0 so
* that the buffer is NOT cancelled.
*/
ASSERT(!(flags & XFS_BLF_CANCEL));
return NULL;
}
/*
* If the buffer is being cancelled then return 1 so that it will be cancelled,
* otherwise return 0. If the buffer is actually a buffer cancel item
* (XFS_BLF_CANCEL is set), then decrement the refcount on the entry in the
* table and remove it from the table if this is the last reference.
*
* We remove the cancel record from the table when we encounter its last
* occurrence in the log so that if the same buffer is re-used again after its
* last cancellation we actually replay the changes made at that point.
*/
STATIC int
xlog_check_buffer_cancelled(
struct xlog *log,
xfs_daddr_t blkno,
uint len,
unsigned short flags)
{
struct xfs_buf_cancel *bcp;
bcp = xlog_peek_buffer_cancelled(log, blkno, len, flags);
if (!bcp)
return 0;
/*
* We've go a match, so return 1 so that the recovery of this buffer
* is cancelled. If this buffer is actually a buffer cancel log
* item, then decrement the refcount on the one in the table and
* remove it if this is the last reference.
*/
if (flags & XFS_BLF_CANCEL) {
if (--bcp->bc_refcount == 0) {
list_del(&bcp->bc_list);
kmem_free(bcp);
}
}
return 1;
}
/*
* Perform recovery for a buffer full of inodes. In these buffers, the only
* data which should be recovered is that which corresponds to the
* di_next_unlinked pointers in the on disk inode structures. The rest of the
* data for the inodes is always logged through the inodes themselves rather
* than the inode buffer and is recovered in xlog_recover_inode_pass2().
*
* The only time when buffers full of inodes are fully recovered is when the
* buffer is full of newly allocated inodes. In this case the buffer will
* not be marked as an inode buffer and so will be sent to
* xlog_recover_do_reg_buffer() below during recovery.
*/
STATIC int
xlog_recover_do_inode_buffer(
struct xfs_mount *mp,
xlog_recover_item_t *item,
struct xfs_buf *bp,
xfs_buf_log_format_t *buf_f)
{
int i;
int item_index = 0;
int bit = 0;
int nbits = 0;
int reg_buf_offset = 0;
int reg_buf_bytes = 0;
int next_unlinked_offset;
int inodes_per_buf;
xfs_agino_t *logged_nextp;
xfs_agino_t *buffer_nextp;
trace_xfs_log_recover_buf_inode_buf(mp->m_log, buf_f);
/*
* Post recovery validation only works properly on CRC enabled
* filesystems.
*/
if (xfs_sb_version_hascrc(&mp->m_sb))
bp->b_ops = &xfs_inode_buf_ops;
inodes_per_buf = BBTOB(bp->b_io_length) >> mp->m_sb.sb_inodelog;
for (i = 0; i < inodes_per_buf; i++) {
next_unlinked_offset = (i * mp->m_sb.sb_inodesize) +
offsetof(xfs_dinode_t, di_next_unlinked);
while (next_unlinked_offset >=
(reg_buf_offset + reg_buf_bytes)) {
/*
* The next di_next_unlinked field is beyond
* the current logged region. Find the next
* logged region that contains or is beyond
* the current di_next_unlinked field.
*/
bit += nbits;
bit = xfs_next_bit(buf_f->blf_data_map,
buf_f->blf_map_size, bit);
/*
* If there are no more logged regions in the
* buffer, then we're done.
*/
if (bit == -1)
return 0;
nbits = xfs_contig_bits(buf_f->blf_data_map,
buf_f->blf_map_size, bit);
ASSERT(nbits > 0);
reg_buf_offset = bit << XFS_BLF_SHIFT;
reg_buf_bytes = nbits << XFS_BLF_SHIFT;
item_index++;
}
/*
* If the current logged region starts after the current
* di_next_unlinked field, then move on to the next
* di_next_unlinked field.
*/
if (next_unlinked_offset < reg_buf_offset)
continue;
ASSERT(item->ri_buf[item_index].i_addr != NULL);
ASSERT((item->ri_buf[item_index].i_len % XFS_BLF_CHUNK) == 0);
ASSERT((reg_buf_offset + reg_buf_bytes) <=
BBTOB(bp->b_io_length));
/*
* The current logged region contains a copy of the
* current di_next_unlinked field. Extract its value
* and copy it to the buffer copy.
*/
logged_nextp = item->ri_buf[item_index].i_addr +
next_unlinked_offset - reg_buf_offset;
if (unlikely(*logged_nextp == 0)) {
xfs_alert(mp,
"Bad inode buffer log record (ptr = 0x%p, bp = 0x%p). "
"Trying to replay bad (0) inode di_next_unlinked field.",
item, bp);
XFS_ERROR_REPORT("xlog_recover_do_inode_buf",
XFS_ERRLEVEL_LOW, mp);
return -EFSCORRUPTED;
}
buffer_nextp = xfs_buf_offset(bp, next_unlinked_offset);
*buffer_nextp = *logged_nextp;
/*
* If necessary, recalculate the CRC in the on-disk inode. We
* have to leave the inode in a consistent state for whoever
* reads it next....
*/
xfs_dinode_calc_crc(mp,
xfs_buf_offset(bp, i * mp->m_sb.sb_inodesize));
}
return 0;
}
/*
* V5 filesystems know the age of the buffer on disk being recovered. We can
* have newer objects on disk than we are replaying, and so for these cases we
* don't want to replay the current change as that will make the buffer contents
* temporarily invalid on disk.
*
* The magic number might not match the buffer type we are going to recover
* (e.g. reallocated blocks), so we ignore the xfs_buf_log_format flags. Hence
* extract the LSN of the existing object in the buffer based on it's current
* magic number. If we don't recognise the magic number in the buffer, then
* return a LSN of -1 so that the caller knows it was an unrecognised block and
* so can recover the buffer.
*
* Note: we cannot rely solely on magic number matches to determine that the
* buffer has a valid LSN - we also need to verify that it belongs to this
* filesystem, so we need to extract the object's LSN and compare it to that
* which we read from the superblock. If the UUIDs don't match, then we've got a
* stale metadata block from an old filesystem instance that we need to recover
* over the top of.
*/
static xfs_lsn_t
xlog_recover_get_buf_lsn(
struct xfs_mount *mp,
struct xfs_buf *bp)
{
__uint32_t magic32;
__uint16_t magic16;
__uint16_t magicda;
void *blk = bp->b_addr;
uuid_t *uuid;
xfs_lsn_t lsn = -1;
/* v4 filesystems always recover immediately */
if (!xfs_sb_version_hascrc(&mp->m_sb))
goto recover_immediately;
magic32 = be32_to_cpu(*(__be32 *)blk);
switch (magic32) {
case XFS_ABTB_CRC_MAGIC:
case XFS_ABTC_CRC_MAGIC:
case XFS_ABTB_MAGIC:
case XFS_ABTC_MAGIC:
case XFS_RMAP_CRC_MAGIC:
case XFS_REFC_CRC_MAGIC:
case XFS_IBT_CRC_MAGIC:
case XFS_IBT_MAGIC: {
struct xfs_btree_block *btb = blk;
lsn = be64_to_cpu(btb->bb_u.s.bb_lsn);
uuid = &btb->bb_u.s.bb_uuid;
break;
}
case XFS_BMAP_CRC_MAGIC:
case XFS_BMAP_MAGIC: {
struct xfs_btree_block *btb = blk;
lsn = be64_to_cpu(btb->bb_u.l.bb_lsn);
uuid = &btb->bb_u.l.bb_uuid;
break;
}
case XFS_AGF_MAGIC:
lsn = be64_to_cpu(((struct xfs_agf *)blk)->agf_lsn);
uuid = &((struct xfs_agf *)blk)->agf_uuid;
break;
case XFS_AGFL_MAGIC:
lsn = be64_to_cpu(((struct xfs_agfl *)blk)->agfl_lsn);
uuid = &((struct xfs_agfl *)blk)->agfl_uuid;
break;
case XFS_AGI_MAGIC:
lsn = be64_to_cpu(((struct xfs_agi *)blk)->agi_lsn);
uuid = &((struct xfs_agi *)blk)->agi_uuid;
break;
case XFS_SYMLINK_MAGIC:
lsn = be64_to_cpu(((struct xfs_dsymlink_hdr *)blk)->sl_lsn);
uuid = &((struct xfs_dsymlink_hdr *)blk)->sl_uuid;
break;
case XFS_DIR3_BLOCK_MAGIC:
case XFS_DIR3_DATA_MAGIC:
case XFS_DIR3_FREE_MAGIC:
lsn = be64_to_cpu(((struct xfs_dir3_blk_hdr *)blk)->lsn);
uuid = &((struct xfs_dir3_blk_hdr *)blk)->uuid;
break;
case XFS_ATTR3_RMT_MAGIC:
/*
* Remote attr blocks are written synchronously, rather than
* being logged. That means they do not contain a valid LSN
* (i.e. transactionally ordered) in them, and hence any time we
* see a buffer to replay over the top of a remote attribute
* block we should simply do so.
*/
goto recover_immediately;
case XFS_SB_MAGIC:
/*
* superblock uuids are magic. We may or may not have a
* sb_meta_uuid on disk, but it will be set in the in-core
* superblock. We set the uuid pointer for verification
* according to the superblock feature mask to ensure we check
* the relevant UUID in the superblock.
*/
lsn = be64_to_cpu(((struct xfs_dsb *)blk)->sb_lsn);
if (xfs_sb_version_hasmetauuid(&mp->m_sb))
uuid = &((struct xfs_dsb *)blk)->sb_meta_uuid;
else
uuid = &((struct xfs_dsb *)blk)->sb_uuid;
break;
default:
break;
}
if (lsn != (xfs_lsn_t)-1) {
if (!uuid_equal(&mp->m_sb.sb_meta_uuid, uuid))
goto recover_immediately;
return lsn;
}
magicda = be16_to_cpu(((struct xfs_da_blkinfo *)blk)->magic);
switch (magicda) {
case XFS_DIR3_LEAF1_MAGIC:
case XFS_DIR3_LEAFN_MAGIC:
case XFS_DA3_NODE_MAGIC:
lsn = be64_to_cpu(((struct xfs_da3_blkinfo *)blk)->lsn);
uuid = &((struct xfs_da3_blkinfo *)blk)->uuid;
break;
default:
break;
}
if (lsn != (xfs_lsn_t)-1) {
if (!uuid_equal(&mp->m_sb.sb_uuid, uuid))
goto recover_immediately;
return lsn;
}
/*
* We do individual object checks on dquot and inode buffers as they
* have their own individual LSN records. Also, we could have a stale
* buffer here, so we have to at least recognise these buffer types.
*
* A notd complexity here is inode unlinked list processing - it logs
* the inode directly in the buffer, but we don't know which inodes have
* been modified, and there is no global buffer LSN. Hence we need to
* recover all inode buffer types immediately. This problem will be
* fixed by logical logging of the unlinked list modifications.
*/
magic16 = be16_to_cpu(*(__be16 *)blk);
switch (magic16) {
case XFS_DQUOT_MAGIC:
case XFS_DINODE_MAGIC:
goto recover_immediately;
default:
break;
}
/* unknown buffer contents, recover immediately */
recover_immediately:
return (xfs_lsn_t)-1;
}
/*
* Validate the recovered buffer is of the correct type and attach the
* appropriate buffer operations to them for writeback. Magic numbers are in a
* few places:
* the first 16 bits of the buffer (inode buffer, dquot buffer),
* the first 32 bits of the buffer (most blocks),
* inside a struct xfs_da_blkinfo at the start of the buffer.
*/
static void
xlog_recover_validate_buf_type(
struct xfs_mount *mp,
struct xfs_buf *bp,
xfs_buf_log_format_t *buf_f,
xfs_lsn_t current_lsn)
{
struct xfs_da_blkinfo *info = bp->b_addr;
__uint32_t magic32;
__uint16_t magic16;
__uint16_t magicda;
char *warnmsg = NULL;
/*
* We can only do post recovery validation on items on CRC enabled
* fielsystems as we need to know when the buffer was written to be able
* to determine if we should have replayed the item. If we replay old
* metadata over a newer buffer, then it will enter a temporarily
* inconsistent state resulting in verification failures. Hence for now
* just avoid the verification stage for non-crc filesystems
*/
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
magic32 = be32_to_cpu(*(__be32 *)bp->b_addr);
magic16 = be16_to_cpu(*(__be16*)bp->b_addr);
magicda = be16_to_cpu(info->magic);
switch (xfs_blft_from_flags(buf_f)) {
case XFS_BLFT_BTREE_BUF:
switch (magic32) {
case XFS_ABTB_CRC_MAGIC:
case XFS_ABTC_CRC_MAGIC:
case XFS_ABTB_MAGIC:
case XFS_ABTC_MAGIC:
bp->b_ops = &xfs_allocbt_buf_ops;
break;
case XFS_IBT_CRC_MAGIC:
case XFS_FIBT_CRC_MAGIC:
case XFS_IBT_MAGIC:
case XFS_FIBT_MAGIC:
bp->b_ops = &xfs_inobt_buf_ops;
break;
case XFS_BMAP_CRC_MAGIC:
case XFS_BMAP_MAGIC:
bp->b_ops = &xfs_bmbt_buf_ops;
break;
case XFS_RMAP_CRC_MAGIC:
bp->b_ops = &xfs_rmapbt_buf_ops;
break;
case XFS_REFC_CRC_MAGIC:
bp->b_ops = &xfs_refcountbt_buf_ops;
break;
default:
warnmsg = "Bad btree block magic!";
break;
}
break;
case XFS_BLFT_AGF_BUF:
if (magic32 != XFS_AGF_MAGIC) {
warnmsg = "Bad AGF block magic!";
break;
}
bp->b_ops = &xfs_agf_buf_ops;
break;
case XFS_BLFT_AGFL_BUF:
if (magic32 != XFS_AGFL_MAGIC) {
warnmsg = "Bad AGFL block magic!";
break;
}
bp->b_ops = &xfs_agfl_buf_ops;
break;
case XFS_BLFT_AGI_BUF:
if (magic32 != XFS_AGI_MAGIC) {
warnmsg = "Bad AGI block magic!";
break;
}
bp->b_ops = &xfs_agi_buf_ops;
break;
case XFS_BLFT_UDQUOT_BUF:
case XFS_BLFT_PDQUOT_BUF:
case XFS_BLFT_GDQUOT_BUF:
#ifdef CONFIG_XFS_QUOTA
if (magic16 != XFS_DQUOT_MAGIC) {
warnmsg = "Bad DQUOT block magic!";
break;
}
bp->b_ops = &xfs_dquot_buf_ops;
#else
xfs_alert(mp,
"Trying to recover dquots without QUOTA support built in!");
ASSERT(0);
#endif
break;
case XFS_BLFT_DINO_BUF:
if (magic16 != XFS_DINODE_MAGIC) {
warnmsg = "Bad INODE block magic!";
break;
}
bp->b_ops = &xfs_inode_buf_ops;
break;
case XFS_BLFT_SYMLINK_BUF:
if (magic32 != XFS_SYMLINK_MAGIC) {
warnmsg = "Bad symlink block magic!";
break;
}
bp->b_ops = &xfs_symlink_buf_ops;
break;
case XFS_BLFT_DIR_BLOCK_BUF:
if (magic32 != XFS_DIR2_BLOCK_MAGIC &&
magic32 != XFS_DIR3_BLOCK_MAGIC) {
warnmsg = "Bad dir block magic!";
break;
}
bp->b_ops = &xfs_dir3_block_buf_ops;
break;
case XFS_BLFT_DIR_DATA_BUF:
if (magic32 != XFS_DIR2_DATA_MAGIC &&
magic32 != XFS_DIR3_DATA_MAGIC) {
warnmsg = "Bad dir data magic!";
break;
}
bp->b_ops = &xfs_dir3_data_buf_ops;
break;
case XFS_BLFT_DIR_FREE_BUF:
if (magic32 != XFS_DIR2_FREE_MAGIC &&
magic32 != XFS_DIR3_FREE_MAGIC) {
warnmsg = "Bad dir3 free magic!";
break;
}
bp->b_ops = &xfs_dir3_free_buf_ops;
break;
case XFS_BLFT_DIR_LEAF1_BUF:
if (magicda != XFS_DIR2_LEAF1_MAGIC &&
magicda != XFS_DIR3_LEAF1_MAGIC) {
warnmsg = "Bad dir leaf1 magic!";
break;
}
bp->b_ops = &xfs_dir3_leaf1_buf_ops;
break;
case XFS_BLFT_DIR_LEAFN_BUF:
if (magicda != XFS_DIR2_LEAFN_MAGIC &&
magicda != XFS_DIR3_LEAFN_MAGIC) {
warnmsg = "Bad dir leafn magic!";
break;
}
bp->b_ops = &xfs_dir3_leafn_buf_ops;
break;
case XFS_BLFT_DA_NODE_BUF:
if (magicda != XFS_DA_NODE_MAGIC &&
magicda != XFS_DA3_NODE_MAGIC) {
warnmsg = "Bad da node magic!";
break;
}
bp->b_ops = &xfs_da3_node_buf_ops;
break;
case XFS_BLFT_ATTR_LEAF_BUF:
if (magicda != XFS_ATTR_LEAF_MAGIC &&
magicda != XFS_ATTR3_LEAF_MAGIC) {
warnmsg = "Bad attr leaf magic!";
break;
}
bp->b_ops = &xfs_attr3_leaf_buf_ops;
break;
case XFS_BLFT_ATTR_RMT_BUF:
if (magic32 != XFS_ATTR3_RMT_MAGIC) {
warnmsg = "Bad attr remote magic!";
break;
}
bp->b_ops = &xfs_attr3_rmt_buf_ops;
break;
case XFS_BLFT_SB_BUF:
if (magic32 != XFS_SB_MAGIC) {
warnmsg = "Bad SB block magic!";
break;
}
bp->b_ops = &xfs_sb_buf_ops;
break;
#ifdef CONFIG_XFS_RT
case XFS_BLFT_RTBITMAP_BUF:
case XFS_BLFT_RTSUMMARY_BUF:
/* no magic numbers for verification of RT buffers */
bp->b_ops = &xfs_rtbuf_ops;
break;
#endif /* CONFIG_XFS_RT */
default:
xfs_warn(mp, "Unknown buffer type %d!",
xfs_blft_from_flags(buf_f));
break;
}
/*
* Nothing else to do in the case of a NULL current LSN as this means
* the buffer is more recent than the change in the log and will be
* skipped.
*/
if (current_lsn == NULLCOMMITLSN)
return;
if (warnmsg) {
xfs_warn(mp, warnmsg);
ASSERT(0);
}
/*
* We must update the metadata LSN of the buffer as it is written out to
* ensure that older transactions never replay over this one and corrupt
* the buffer. This can occur if log recovery is interrupted at some
* point after the current transaction completes, at which point a
* subsequent mount starts recovery from the beginning.
*
* Write verifiers update the metadata LSN from log items attached to
* the buffer. Therefore, initialize a bli purely to carry the LSN to
* the verifier. We'll clean it up in our ->iodone() callback.
*/
if (bp->b_ops) {
struct xfs_buf_log_item *bip;
ASSERT(!bp->b_iodone || bp->b_iodone == xlog_recover_iodone);
bp->b_iodone = xlog_recover_iodone;
xfs_buf_item_init(bp, mp);
bip = bp->b_fspriv;
bip->bli_item.li_lsn = current_lsn;
}
}
/*
* Perform a 'normal' buffer recovery. Each logged region of the
* buffer should be copied over the corresponding region in the
* given buffer. The bitmap in the buf log format structure indicates
* where to place the logged data.
*/
STATIC void
xlog_recover_do_reg_buffer(
struct xfs_mount *mp,
xlog_recover_item_t *item,
struct xfs_buf *bp,
xfs_buf_log_format_t *buf_f,
xfs_lsn_t current_lsn)
{
int i;
int bit;
int nbits;
int error;
trace_xfs_log_recover_buf_reg_buf(mp->m_log, buf_f);
bit = 0;
i = 1; /* 0 is the buf format structure */
while (1) {
bit = xfs_next_bit(buf_f->blf_data_map,
buf_f->blf_map_size, bit);
if (bit == -1)
break;
nbits = xfs_contig_bits(buf_f->blf_data_map,
buf_f->blf_map_size, bit);
ASSERT(nbits > 0);
ASSERT(item->ri_buf[i].i_addr != NULL);
ASSERT(item->ri_buf[i].i_len % XFS_BLF_CHUNK == 0);
ASSERT(BBTOB(bp->b_io_length) >=
((uint)bit << XFS_BLF_SHIFT) + (nbits << XFS_BLF_SHIFT));
/*
* The dirty regions logged in the buffer, even though
* contiguous, may span multiple chunks. This is because the
* dirty region may span a physical page boundary in a buffer
* and hence be split into two separate vectors for writing into
* the log. Hence we need to trim nbits back to the length of
* the current region being copied out of the log.
*/
if (item->ri_buf[i].i_len < (nbits << XFS_BLF_SHIFT))
nbits = item->ri_buf[i].i_len >> XFS_BLF_SHIFT;
/*
* Do a sanity check if this is a dquot buffer. Just checking
* the first dquot in the buffer should do. XXXThis is
* probably a good thing to do for other buf types also.
*/
error = 0;
if (buf_f->blf_flags &
(XFS_BLF_UDQUOT_BUF|XFS_BLF_PDQUOT_BUF|XFS_BLF_GDQUOT_BUF)) {
if (item->ri_buf[i].i_addr == NULL) {
xfs_alert(mp,
"XFS: NULL dquot in %s.", __func__);
goto next;
}
if (item->ri_buf[i].i_len < sizeof(xfs_disk_dquot_t)) {
xfs_alert(mp,
"XFS: dquot too small (%d) in %s.",
item->ri_buf[i].i_len, __func__);
goto next;
}
error = xfs_dqcheck(mp, item->ri_buf[i].i_addr,
-1, 0, XFS_QMOPT_DOWARN,
"dquot_buf_recover");
if (error)
goto next;
}
memcpy(xfs_buf_offset(bp,
(uint)bit << XFS_BLF_SHIFT), /* dest */
item->ri_buf[i].i_addr, /* source */
nbits<<XFS_BLF_SHIFT); /* length */
next:
i++;
bit += nbits;
}
/* Shouldn't be any more regions */
ASSERT(i == item->ri_total);
xlog_recover_validate_buf_type(mp, bp, buf_f, current_lsn);
}
/*
* Perform a dquot buffer recovery.
* Simple algorithm: if we have found a QUOTAOFF log item of the same type
* (ie. USR or GRP), then just toss this buffer away; don't recover it.
* Else, treat it as a regular buffer and do recovery.
*
* Return false if the buffer was tossed and true if we recovered the buffer to
* indicate to the caller if the buffer needs writing.
*/
STATIC bool
xlog_recover_do_dquot_buffer(
struct xfs_mount *mp,
struct xlog *log,
struct xlog_recover_item *item,
struct xfs_buf *bp,
struct xfs_buf_log_format *buf_f)
{
uint type;
trace_xfs_log_recover_buf_dquot_buf(log, buf_f);
/*
* Filesystems are required to send in quota flags at mount time.
*/
if (!mp->m_qflags)
return false;
type = 0;
if (buf_f->blf_flags & XFS_BLF_UDQUOT_BUF)
type |= XFS_DQ_USER;
if (buf_f->blf_flags & XFS_BLF_PDQUOT_BUF)
type |= XFS_DQ_PROJ;
if (buf_f->blf_flags & XFS_BLF_GDQUOT_BUF)
type |= XFS_DQ_GROUP;
/*
* This type of quotas was turned off, so ignore this buffer
*/
if (log->l_quotaoffs_flag & type)
return false;
xlog_recover_do_reg_buffer(mp, item, bp, buf_f, NULLCOMMITLSN);
return true;
}
/*
* This routine replays a modification made to a buffer at runtime.
* There are actually two types of buffer, regular and inode, which
* are handled differently. Inode buffers are handled differently
* in that we only recover a specific set of data from them, namely
* the inode di_next_unlinked fields. This is because all other inode
* data is actually logged via inode records and any data we replay
* here which overlaps that may be stale.
*
* When meta-data buffers are freed at run time we log a buffer item
* with the XFS_BLF_CANCEL bit set to indicate that previous copies
* of the buffer in the log should not be replayed at recovery time.
* This is so that if the blocks covered by the buffer are reused for
* file data before we crash we don't end up replaying old, freed
* meta-data into a user's file.
*
* To handle the cancellation of buffer log items, we make two passes
* over the log during recovery. During the first we build a table of
* those buffers which have been cancelled, and during the second we
* only replay those buffers which do not have corresponding cancel
* records in the table. See xlog_recover_buffer_pass[1,2] above
* for more details on the implementation of the table of cancel records.
*/
STATIC int
xlog_recover_buffer_pass2(
struct xlog *log,
struct list_head *buffer_list,
struct xlog_recover_item *item,
xfs_lsn_t current_lsn)
{
xfs_buf_log_format_t *buf_f = item->ri_buf[0].i_addr;
xfs_mount_t *mp = log->l_mp;
xfs_buf_t *bp;
int error;
uint buf_flags;
xfs_lsn_t lsn;
/*
* In this pass we only want to recover all the buffers which have
* not been cancelled and are not cancellation buffers themselves.
*/
if (xlog_check_buffer_cancelled(log, buf_f->blf_blkno,
buf_f->blf_len, buf_f->blf_flags)) {
trace_xfs_log_recover_buf_cancel(log, buf_f);
return 0;
}
trace_xfs_log_recover_buf_recover(log, buf_f);
buf_flags = 0;
if (buf_f->blf_flags & XFS_BLF_INODE_BUF)
buf_flags |= XBF_UNMAPPED;
bp = xfs_buf_read(mp->m_ddev_targp, buf_f->blf_blkno, buf_f->blf_len,
buf_flags, NULL);
if (!bp)
return -ENOMEM;
error = bp->b_error;
if (error) {
xfs_buf_ioerror_alert(bp, "xlog_recover_do..(read#1)");
goto out_release;
}
/*
* Recover the buffer only if we get an LSN from it and it's less than
* the lsn of the transaction we are replaying.
*
* Note that we have to be extremely careful of readahead here.
* Readahead does not attach verfiers to the buffers so if we don't
* actually do any replay after readahead because of the LSN we found
* in the buffer if more recent than that current transaction then we
* need to attach the verifier directly. Failure to do so can lead to
* future recovery actions (e.g. EFI and unlinked list recovery) can
* operate on the buffers and they won't get the verifier attached. This
* can lead to blocks on disk having the correct content but a stale
* CRC.
*
* It is safe to assume these clean buffers are currently up to date.
* If the buffer is dirtied by a later transaction being replayed, then
* the verifier will be reset to match whatever recover turns that
* buffer into.
*/
lsn = xlog_recover_get_buf_lsn(mp, bp);
if (lsn && lsn != -1 && XFS_LSN_CMP(lsn, current_lsn) >= 0) {
trace_xfs_log_recover_buf_skip(log, buf_f);
xlog_recover_validate_buf_type(mp, bp, buf_f, NULLCOMMITLSN);
goto out_release;
}
if (buf_f->blf_flags & XFS_BLF_INODE_BUF) {
error = xlog_recover_do_inode_buffer(mp, item, bp, buf_f);
if (error)
goto out_release;
} else if (buf_f->blf_flags &
(XFS_BLF_UDQUOT_BUF|XFS_BLF_PDQUOT_BUF|XFS_BLF_GDQUOT_BUF)) {
bool dirty;
dirty = xlog_recover_do_dquot_buffer(mp, log, item, bp, buf_f);
if (!dirty)
goto out_release;
} else {
xlog_recover_do_reg_buffer(mp, item, bp, buf_f, current_lsn);
}
/*
* Perform delayed write on the buffer. Asynchronous writes will be
* slower when taking into account all the buffers to be flushed.
*
* Also make sure that only inode buffers with good sizes stay in
* the buffer cache. The kernel moves inodes in buffers of 1 block
* or mp->m_inode_cluster_size bytes, whichever is bigger. The inode
* buffers in the log can be a different size if the log was generated
* by an older kernel using unclustered inode buffers or a newer kernel
* running with a different inode cluster size. Regardless, if the
* the inode buffer size isn't MAX(blocksize, mp->m_inode_cluster_size)
* for *our* value of mp->m_inode_cluster_size, then we need to keep
* the buffer out of the buffer cache so that the buffer won't
* overlap with future reads of those inodes.
*/
if (XFS_DINODE_MAGIC ==
be16_to_cpu(*((__be16 *)xfs_buf_offset(bp, 0))) &&
(BBTOB(bp->b_io_length) != MAX(log->l_mp->m_sb.sb_blocksize,
(__uint32_t)log->l_mp->m_inode_cluster_size))) {
xfs_buf_stale(bp);
error = xfs_bwrite(bp);
} else {
ASSERT(bp->b_target->bt_mount == mp);
bp->b_iodone = xlog_recover_iodone;
xfs_buf_delwri_queue(bp, buffer_list);
}
out_release:
xfs_buf_relse(bp);
return error;
}
/*
* Inode fork owner changes
*
* If we have been told that we have to reparent the inode fork, it's because an
* extent swap operation on a CRC enabled filesystem has been done and we are
* replaying it. We need to walk the BMBT of the appropriate fork and change the
* owners of it.
*
* The complexity here is that we don't have an inode context to work with, so
* after we've replayed the inode we need to instantiate one. This is where the
* fun begins.
*
* We are in the middle of log recovery, so we can't run transactions. That
* means we cannot use cache coherent inode instantiation via xfs_iget(), as
* that will result in the corresponding iput() running the inode through
* xfs_inactive(). If we've just replayed an inode core that changes the link
* count to zero (i.e. it's been unlinked), then xfs_inactive() will run
* transactions (bad!).
*
* So, to avoid this, we instantiate an inode directly from the inode core we've
* just recovered. We have the buffer still locked, and all we really need to
* instantiate is the inode core and the forks being modified. We can do this
* manually, then run the inode btree owner change, and then tear down the
* xfs_inode without having to run any transactions at all.
*
* Also, because we don't have a transaction context available here but need to
* gather all the buffers we modify for writeback so we pass the buffer_list
* instead for the operation to use.
*/
STATIC int
xfs_recover_inode_owner_change(
struct xfs_mount *mp,
struct xfs_dinode *dip,
struct xfs_inode_log_format *in_f,
struct list_head *buffer_list)
{
struct xfs_inode *ip;
int error;
ASSERT(in_f->ilf_fields & (XFS_ILOG_DOWNER|XFS_ILOG_AOWNER));
ip = xfs_inode_alloc(mp, in_f->ilf_ino);
if (!ip)
return -ENOMEM;
/* instantiate the inode */
xfs_inode_from_disk(ip, dip);
ASSERT(ip->i_d.di_version >= 3);
error = xfs_iformat_fork(ip, dip);
if (error)
goto out_free_ip;
if (in_f->ilf_fields & XFS_ILOG_DOWNER) {
ASSERT(in_f->ilf_fields & XFS_ILOG_DBROOT);
error = xfs_bmbt_change_owner(NULL, ip, XFS_DATA_FORK,
ip->i_ino, buffer_list);
if (error)
goto out_free_ip;
}
if (in_f->ilf_fields & XFS_ILOG_AOWNER) {
ASSERT(in_f->ilf_fields & XFS_ILOG_ABROOT);
error = xfs_bmbt_change_owner(NULL, ip, XFS_ATTR_FORK,
ip->i_ino, buffer_list);
if (error)
goto out_free_ip;
}
out_free_ip:
xfs_inode_free(ip);
return error;
}
STATIC int
xlog_recover_inode_pass2(
struct xlog *log,
struct list_head *buffer_list,
struct xlog_recover_item *item,
xfs_lsn_t current_lsn)
{
xfs_inode_log_format_t *in_f;
xfs_mount_t *mp = log->l_mp;
xfs_buf_t *bp;
xfs_dinode_t *dip;
int len;