struct ext2_qcow2_hdr *qcow2_read_header(int fd)
{
void *buffer = NULL;
struct ext2_qcow2_hdr *hdr = NULL;
size_t size;
errcode_t ret;
ret = ext2fs_get_mem(sizeof(struct ext2_qcow2_hdr), &buffer);
if (ret)
return NULL;
memset(buffer, 0, sizeof(struct ext2_qcow2_hdr));
if (ext2fs_llseek(fd, 0, SEEK_SET < 0))
return NULL;
size = read(fd, buffer, sizeof(struct ext2_qcow2_hdr));
if (size != sizeof(struct ext2_qcow2_hdr)) {
ext2fs_free_mem(&buffer);
return NULL;
}
hdr = (struct ext2_qcow2_hdr *)(buffer);
if ((ext2fs_be32_to_cpu(hdr->magic) != QCOW_MAGIC) ||
(ext2fs_be32_to_cpu(hdr->version) != 2)) {
ext2fs_free_mem(&hdr);
return NULL;
}
return hdr;
}
static int scan_revoke_records(journal_t *journal, struct buffer_head *bh,
tid_t sequence, struct recovery_info *info)
{
journal_revoke_header_t *header;
int offset, max;
int record_len = 4;
header = (journal_revoke_header_t *) bh->b_data;
offset = sizeof(journal_revoke_header_t);
max = ext2fs_be32_to_cpu(header->r_count);
if (!jbd2_revoke_block_csum_verify(journal, header))
return -EINVAL;
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT))
record_len = 8;
while (offset + record_len <= max) {
unsigned long long blocknr;
int err;
if (record_len == 4)
blocknr = ext2fs_be32_to_cpu(* ((__u32 *) (bh->b_data+offset)));
else
blocknr = ext2fs_be64_to_cpu(* ((__u64 *) (bh->b_data+offset)));
offset += record_len;
err = journal_set_revoke(journal, blocknr, sequence);
if (err)
return err;
++info->nr_revokes;
}
return 0;
}
static inline unsigned long long read_tag_block(journal_t *journal,
journal_block_tag_t *tag)
{
unsigned long long block = ext2fs_be32_to_cpu(tag->t_blocknr);
if (jfs_has_feature_64bit(journal))
block |= (u64)ext2fs_be32_to_cpu(tag->t_blocknr_high) << 32;
return block;
}
static inline unsigned long long read_tag_block(journal_t *journal,
journal_block_tag_t *tag)
{
unsigned long long block = ext2fs_be32_to_cpu(tag->t_blocknr);
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT))
block |= (u64)ext2fs_be32_to_cpu(tag->t_blocknr_high) << 32;
return block;
}
/**
* journal_recover - recovers a on-disk journal
* @journal: the journal to recover
*
* The primary function for recovering the log contents when mounting a
* journaled device.
*
* Recovery is done in three passes. In the first pass, we look for the
* end of the log. In the second, we assemble the list of revoke
* blocks. In the third and final pass, we replay any un-revoked blocks
* in the log.
*/
int journal_recover(journal_t *journal)
{
int err, err2;
journal_superblock_t * sb;
struct recovery_info info;
memset(&info, 0, sizeof(info));
sb = journal->j_superblock;
/*
* The journal superblock's s_start field (the current log head)
* is always zero if, and only if, the journal was cleanly
* unmounted.
*/
if (!sb->s_start) {
jbd_debug(1, "No recovery required, last transaction %d\n",
ext2fs_be32_to_cpu(sb->s_sequence));
journal->j_transaction_sequence = ext2fs_be32_to_cpu(sb->s_sequence) + 1;
return 0;
}
err = do_one_pass(journal, &info, PASS_SCAN);
if (!err)
err = do_one_pass(journal, &info, PASS_REVOKE);
if (!err)
err = do_one_pass(journal, &info, PASS_REPLAY);
jbd_debug(1, "JBD2: recovery, exit status %d, "
"recovered transactions %u to %u\n",
err, info.start_transaction, info.end_transaction);
jbd_debug(1, "JBD2: Replayed %d and revoked %d/%d blocks\n",
info.nr_replays, info.nr_revoke_hits, info.nr_revokes);
/* Restart the log at the next transaction ID, thus invalidating
* any existing commit records in the log. */
journal->j_transaction_sequence = ++info.end_transaction;
journal_clear_revoke(journal);
err2 = sync_blockdev(journal->j_fs_dev);
if (!err)
err = err2;
/* Make sure all replayed data is on permanent storage */
if (journal->j_flags & JFS_BARRIER) {
err2 = blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL);
if (!err)
err = err2;
}
return err;
}
/**
* journal_skip_recovery - Start journal and wipe exiting records
* @journal: journal to startup
*
* Locate any valid recovery information from the journal and set up the
* journal structures in memory to ignore it (presumably because the
* caller has evidence that it is out of date).
* This function does'nt appear to be exorted..
*
* We perform one pass over the journal to allow us to tell the user how
* much recovery information is being erased, and to let us initialise
* the journal transaction sequence numbers to the next unused ID.
*/
int journal_skip_recovery(journal_t *journal)
{
int err;
struct recovery_info info;
memset (&info, 0, sizeof(info));
err = do_one_pass(journal, &info, PASS_SCAN);
if (err) {
printk(KERN_ERR "JBD2: error %d scanning journal\n", err);
++journal->j_transaction_sequence;
} else {
#ifdef CONFIG_JFS_DEBUG
int dropped = info.end_transaction -
ext2fs_be32_to_cpu(journal->j_superblock->s_sequence);
jbd_debug(1,
"JBD2: ignoring %d transaction%s from the journal.\n",
dropped, (dropped == 1) ? "" : "s");
#endif
journal->j_transaction_sequence = ++info.end_transaction;
}
journal->j_tail = 0;
return err;
}
static int scan_revoke_records(journal_t *journal, struct buffer_head *bh,
tid_t sequence, struct recovery_info *info)
{
journal_revoke_header_t *header;
int offset, max;
int csum_size = 0;
__u32 rcount;
int record_len = 4;
header = (journal_revoke_header_t *) bh->b_data;
offset = sizeof(journal_revoke_header_t);
rcount = ext2fs_be32_to_cpu(header->r_count);
if (!jbd2_revoke_block_csum_verify(journal, header))
return -EFSBADCRC;
if (journal_has_csum_v2or3(journal))
csum_size = sizeof(struct journal_revoke_tail);
if (rcount > journal->j_blocksize - csum_size)
return -EINVAL;
max = rcount;
if (jfs_has_feature_64bit(journal))
record_len = 8;
while (offset + record_len <= max) {
unsigned long long blocknr;
int err;
if (record_len == 4)
blocknr = ext2fs_be32_to_cpu(* ((__u32 *) (bh->b_data+offset)));
else
blocknr = ext2fs_be64_to_cpu(* ((__u64 *) (bh->b_data+offset)));
offset += record_len;
err = journal_set_revoke(journal, blocknr, sequence);
if (err)
return err;
++info->nr_revokes;
}
return 0;
}
static void print_journal_information(ext2_filsys fs)
{
errcode_t retval;
char buf[1024];
char str[80];
unsigned int i;
journal_superblock_t *jsb;
/* Get the journal superblock */
if ((retval = io_channel_read_blk64(fs->io,
fs->super->s_first_data_block + 1,
-1024, buf))) {
com_err(program_name, retval, "%s",
_("while reading journal superblock"));
exit(1);
}
jsb = (journal_superblock_t *) buf;
if ((jsb->s_header.h_magic != (unsigned) ntohl(JFS_MAGIC_NUMBER)) ||
(jsb->s_header.h_blocktype !=
(unsigned) ntohl(JFS_SUPERBLOCK_V2))) {
com_err(program_name, 0, "%s",
_("Couldn't find journal superblock magic numbers"));
exit(1);
}
if (jsb->s_feature_compat &
ext2fs_cpu_to_be32(JFS_FEATURE_COMPAT_CHECKSUM))
printf(_("Journal checksum type: crc32\n"));
if (jsb->s_feature_incompat &
ext2fs_cpu_to_be32(JFS_FEATURE_INCOMPAT_CSUM_V2))
printf(_("Journal checksum type: %s\n"
"Journal checksum: 0x%08x\n"),
journal_checksum_type_str(jsb->s_checksum_type),
ext2fs_be32_to_cpu(jsb->s_checksum));
printf(_("\nJournal block size: %u\n"
"Journal length: %u\n"
"Journal first block: %u\n"
"Journal sequence: 0x%08x\n"
"Journal start: %u\n"
"Journal number of users: %u\n"),
(unsigned int)ntohl(jsb->s_blocksize), (unsigned int)ntohl(jsb->s_maxlen),
(unsigned int)ntohl(jsb->s_first), (unsigned int)ntohl(jsb->s_sequence),
(unsigned int)ntohl(jsb->s_start), (unsigned int)ntohl(jsb->s_nr_users));
for (i=0; i < ntohl(jsb->s_nr_users); i++) {
uuid_unparse(&jsb->s_users[i*16], str);
printf(i ? " %s\n"
: _("Journal users: %s\n"),
str);
}
}
static int e2fsck_journal_sb_csum_verify(journal_t *j,
journal_superblock_t *jsb)
{
__u32 provided, calculated;
if (!journal_has_csum_v2or3(j))
return 1;
provided = ext2fs_be32_to_cpu(jsb->s_checksum);
calculated = e2fsck_journal_sb_csum(jsb);
return provided == calculated;
}
static int e2fsck_journal_sb_csum_verify(journal_t *j,
journal_superblock_t *jsb)
{
__u32 provided, calculated;
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 1;
provided = ext2fs_be32_to_cpu(jsb->s_checksum);
calculated = e2fsck_journal_sb_csum(jsb);
return provided == calculated;
}
static int jbd2_commit_block_csum_verify(journal_t *j, void *buf)
{
struct commit_header *h;
__u32 provided, calculated;
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 1;
h = buf;
provided = h->h_chksum[0];
h->h_chksum[0] = 0;
calculated = ext2fs_crc32c_le(~0, j->j_superblock->s_uuid,
sizeof(j->j_superblock->s_uuid));
calculated = ext2fs_crc32c_le(calculated, buf, j->j_blocksize);
h->h_chksum[0] = provided;
provided = ext2fs_be32_to_cpu(provided);
return provided == calculated;
}
/*
* Check whether given quota file is in our format
*/
static int v2_check_file(struct quota_handle *h, int type, int fmt)
{
struct v2_disk_dqheader dqh;
int file_magics[] = INITQMAGICS;
if (fmt != QFMT_VFS_V1)
return 0;
if (!v2_read_header(h, &dqh))
return 0;
if (ext2fs_le32_to_cpu(dqh.dqh_magic) != file_magics[type]) {
if (ext2fs_be32_to_cpu(dqh.dqh_magic) == file_magics[type])
log_err("Your quota file is stored in wrong endianity");
return 0;
}
if (V2_VERSION != ext2fs_le32_to_cpu(dqh.dqh_version))
return 0;
return 1;
}
static int jbd2_revoke_block_csum_verify(journal_t *j,
void *buf)
{
struct journal_revoke_tail *tail;
__u32 provided, calculated;
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 1;
tail = (struct journal_revoke_tail *)(buf + j->j_blocksize -
sizeof(struct journal_revoke_tail));
provided = tail->r_checksum;
tail->r_checksum = 0;
calculated = ext2fs_crc32c_le(~0, j->j_superblock->s_uuid,
sizeof(j->j_superblock->s_uuid));
calculated = ext2fs_crc32c_le(calculated, buf, j->j_blocksize);
tail->r_checksum = provided;
provided = ext2fs_be32_to_cpu(provided);
return provided == calculated;
}
static void print_inline_journal_information(ext2_filsys fs)
{
journal_superblock_t *jsb;
struct ext2_inode inode;
ext2_file_t journal_file;
errcode_t retval;
ino_t ino = fs->super->s_journal_inum;
char buf[1024];
__u32 *mask_ptr, mask, m;
int i, j, size, printed = 0;
if (fs->flags & EXT2_FLAG_IMAGE_FILE)
return;
retval = ext2fs_read_inode(fs, ino, &inode);
if (retval) {
com_err(program_name, retval, "%s",
_("while reading journal inode"));
exit(1);
}
retval = ext2fs_file_open2(fs, ino, &inode, 0, &journal_file);
if (retval) {
com_err(program_name, retval, "%s",
_("while opening journal inode"));
exit(1);
}
retval = ext2fs_file_read(journal_file, buf, sizeof(buf), 0);
if (retval) {
com_err(program_name, retval, "%s",
_("while reading journal super block"));
exit(1);
}
ext2fs_file_close(journal_file);
jsb = (journal_superblock_t *) buf;
if (be32_to_cpu(jsb->s_header.h_magic) != JFS_MAGIC_NUMBER) {
fprintf(stderr, "%s",
_("Journal superblock magic number invalid!\n"));
exit(1);
}
printf("%s", _("Journal features: "));
for (i=0, mask_ptr=&jsb->s_feature_compat; i <3; i++,mask_ptr++) {
mask = be32_to_cpu(*mask_ptr);
for (j=0,m=1; j < 32; j++, m<<=1) {
if (mask & m) {
printf(" %s", e2p_jrnl_feature2string(i, m));
printed++;
}
}
}
if (printed == 0)
printf(" (none)");
printf("\n");
fputs(_("Journal size: "), stdout);
if ((fs->super->s_feature_ro_compat &
EXT4_FEATURE_RO_COMPAT_HUGE_FILE) &&
(inode.i_flags & EXT4_HUGE_FILE_FL))
size = inode.i_blocks / (fs->blocksize / 1024);
else
size = inode.i_blocks >> 1;
if (size < 8192)
printf("%uk\n", size);
else
printf("%uM\n", size >> 10);
printf(_("Journal length: %u\n"
"Journal sequence: 0x%08x\n"
"Journal start: %u\n"),
(unsigned int)ntohl(jsb->s_maxlen),
(unsigned int)ntohl(jsb->s_sequence),
(unsigned int)ntohl(jsb->s_start));
if (jsb->s_feature_compat &
ext2fs_cpu_to_be32(JFS_FEATURE_COMPAT_CHECKSUM))
printf(_("Journal checksum type: crc32\n"));
if (jsb->s_feature_incompat &
ext2fs_cpu_to_be32(JFS_FEATURE_INCOMPAT_CSUM_V2))
printf(_("Journal checksum type: %s\n"
"Journal checksum: 0x%08x\n"),
journal_checksum_type_str(jsb->s_checksum_type),
ext2fs_be32_to_cpu(jsb->s_checksum));
if (jsb->s_errno != 0)
printf(_("Journal errno: %d\n"),
(int) ntohl(jsb->s_errno));
}
int qcow2_write_raw_image(int qcow2_fd, int raw_fd,
struct ext2_qcow2_hdr *hdr)
{
struct ext2_qcow2_image img;
errcode_t ret = 0;
unsigned int l1_index, l2_index;
ext2_off64_t offset;
blk64_t *l1_table, *l2_table = 0;
void *copy_buf = NULL;
size_t size;
if (hdr->crypt_method)
return -QCOW_ENCRYPTED;
img.fd = qcow2_fd;
img.hdr = hdr;
img.l2_cache = NULL;
img.l1_table = NULL;
img.cluster_bits = ext2fs_be32_to_cpu(hdr->cluster_bits);
img.cluster_size = 1 << img.cluster_bits;
img.l1_size = ext2fs_be32_to_cpu(hdr->l1_size);
img.l1_offset = ext2fs_be64_to_cpu(hdr->l1_table_offset);
img.l2_size = 1 << (img.cluster_bits - 3);
img.image_size = ext2fs_be64_to_cpu(hdr->size);
ret = ext2fs_get_memzero(img.cluster_size, &l2_table);
if (ret)
goto out;
ret = ext2fs_get_memzero(1 << img.cluster_bits, ©_buf);
if (ret)
goto out;
if (ext2fs_llseek(raw_fd, 0, SEEK_SET) < 0) {
ret = errno;
goto out;
}
ret = qcow2_read_l1_table(&img);
if (ret)
goto out;
l1_table = img.l1_table;
/* Walk through l1 table */
for (l1_index = 0; l1_index < img.l1_size; l1_index++) {
ext2_off64_t off_out;
offset = ext2fs_be64_to_cpu(l1_table[l1_index]) &
~QCOW_OFLAG_COPIED;
if ((offset > img.image_size) ||
(offset <= 0))
continue;
if (offset & QCOW_OFLAG_COMPRESSED) {
ret = -QCOW_COMPRESSED;
goto out;
}
ret = qcow2_read_l2_table(&img, offset, &l2_table);
if (ret)
break;
/* Walk through l2 table and copy data blocks into raw image */
for (l2_index = 0; l2_index < img.l2_size; l2_index++) {
offset = ext2fs_be64_to_cpu(l2_table[l2_index]) &
~QCOW_OFLAG_COPIED;
if (offset == 0)
continue;
off_out = (l1_index * img.l2_size) +
l2_index;
off_out <<= img.cluster_bits;
ret = qcow2_copy_data(qcow2_fd, raw_fd, offset,
off_out, copy_buf, img.cluster_size);
if (ret)
goto out;
}
}
/* Resize the output image to the filesystem size */
if (ext2fs_llseek(raw_fd, img.image_size - 1, SEEK_SET) < 0)
return errno;
((char *)copy_buf)[0] = 0;
size = write(raw_fd, copy_buf, 1);
if (size != 1)
return errno;
out:
if (copy_buf)
ext2fs_free_mem(©_buf);
if (img.l1_table)
ext2fs_free_mem(&img.l1_table);
if (l2_table)
ext2fs_free_mem(&l2_table);
return ret;
}
static int do_one_pass(journal_t *journal,
struct recovery_info *info, enum passtype pass)
{
unsigned int first_commit_ID, next_commit_ID;
unsigned long next_log_block;
int err, success = 0;
journal_superblock_t * sb;
journal_header_t * tmp;
struct buffer_head * bh;
unsigned int sequence;
int blocktype;
int tag_bytes = journal_tag_bytes(journal);
__u32 crc32_sum = ~0; /* Transactional Checksums */
int descr_csum_size = 0;
int block_error = 0;
/*
* First thing is to establish what we expect to find in the log
* (in terms of transaction IDs), and where (in terms of log
* block offsets): query the superblock.
*/
sb = journal->j_superblock;
next_commit_ID = ext2fs_be32_to_cpu(sb->s_sequence);
next_log_block = ext2fs_be32_to_cpu(sb->s_start);
first_commit_ID = next_commit_ID;
if (pass == PASS_SCAN)
info->start_transaction = first_commit_ID;
jbd_debug(1, "Starting recovery pass %d\n", pass);
/*
* Now we walk through the log, transaction by transaction,
* making sure that each transaction has a commit block in the
* expected place. Each complete transaction gets replayed back
* into the main filesystem.
*/
while (1) {
int flags;
char * tagp;
journal_block_tag_t * tag;
struct buffer_head * obh;
struct buffer_head * nbh;
cond_resched();
/* If we already know where to stop the log traversal,
* check right now that we haven't gone past the end of
* the log. */
if (pass != PASS_SCAN)
if (tid_geq(next_commit_ID, info->end_transaction))
break;
jbd_debug(2, "Scanning for sequence ID %u at %lu/%lu\n",
next_commit_ID, next_log_block, journal->j_last);
/* Skip over each chunk of the transaction looking
* either the next descriptor block or the final commit
* record. */
jbd_debug(3, "JBD2: checking block %ld\n", next_log_block);
err = jread(&bh, journal, next_log_block);
if (err)
goto failed;
next_log_block++;
wrap(journal, next_log_block);
/* What kind of buffer is it?
*
* If it is a descriptor block, check that it has the
* expected sequence number. Otherwise, we're all done
* here. */
tmp = (journal_header_t *)bh->b_data;
if (tmp->h_magic != ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER)) {
brelse(bh);
break;
}
blocktype = ext2fs_be32_to_cpu(tmp->h_blocktype);
sequence = ext2fs_be32_to_cpu(tmp->h_sequence);
jbd_debug(3, "Found magic %d, sequence %d\n",
blocktype, sequence);
if (sequence != next_commit_ID) {
brelse(bh);
break;
}
/* OK, we have a valid descriptor block which matches
* all of the sequence number checks. What are we going
* to do with it? That depends on the pass... */
switch(blocktype) {
case JFS_DESCRIPTOR_BLOCK:
/* Verify checksum first */
if (journal_has_csum_v2or3(journal))
descr_csum_size =
sizeof(struct journal_block_tail);
if (descr_csum_size > 0 &&
!jbd2_descr_block_csum_verify(journal,
bh->b_data)) {
err = -EFSBADCRC;
brelse(bh);
goto failed;
}
/* If it is a valid descriptor block, replay it
* in pass REPLAY; if journal_checksums enabled, then
* calculate checksums in PASS_SCAN, otherwise,
* just skip over the blocks it describes. */
if (pass != PASS_REPLAY) {
if (pass == PASS_SCAN &&
jfs_has_feature_checksum(journal) &&
!info->end_transaction) {
if (calc_chksums(journal, bh,
&next_log_block,
&crc32_sum)) {
put_bh(bh);
break;
}
put_bh(bh);
continue;
}
next_log_block += count_tags(journal, bh);
wrap(journal, next_log_block);
put_bh(bh);
continue;
}
/* A descriptor block: we can now write all of
* the data blocks. Yay, useful work is finally
* getting done here! */
tagp = &bh->b_data[sizeof(journal_header_t)];
while ((tagp - bh->b_data + tag_bytes)
<= journal->j_blocksize - descr_csum_size) {
unsigned long io_block;
tag = (journal_block_tag_t *) tagp;
flags = ext2fs_be16_to_cpu(tag->t_flags);
io_block = next_log_block++;
wrap(journal, next_log_block);
err = jread(&obh, journal, io_block);
if (err) {
/* Recover what we can, but
* report failure at the end. */
success = err;
printk(KERN_ERR
"JBD2: IO error %d recovering "
"block %ld in log\n",
err, io_block);
} else {
unsigned long long blocknr;
J_ASSERT(obh != NULL);
blocknr = read_tag_block(journal,
tag);
/* If the block has been
* revoked, then we're all done
* here. */
if (journal_test_revoke
(journal, blocknr,
next_commit_ID)) {
brelse(obh);
++info->nr_revoke_hits;
goto skip_write;
}
/* Look for block corruption */
if (!jbd2_block_tag_csum_verify(
journal, tag, obh->b_data,
ext2fs_be32_to_cpu(tmp->h_sequence))) {
brelse(obh);
success = -EFSBADCRC;
printk(KERN_ERR "JBD2: Invalid "
"checksum recovering "
"block %llu in log\n",
blocknr);
block_error = 1;
goto skip_write;
}
/* Find a buffer for the new
* data being restored */
nbh = __getblk(journal->j_fs_dev,
blocknr,
journal->j_blocksize);
if (nbh == NULL) {
printk(KERN_ERR
"JBD2: Out of memory "
"during recovery.\n");
err = -ENOMEM;
brelse(bh);
brelse(obh);
goto failed;
}
lock_buffer(nbh);
memcpy(nbh->b_data, obh->b_data,
journal->j_blocksize);
if (flags & JFS_FLAG_ESCAPE) {
*((__u32 *)nbh->b_data) =
ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER);
}
BUFFER_TRACE(nbh, "marking dirty");
set_buffer_uptodate(nbh);
mark_buffer_dirty(nbh);
BUFFER_TRACE(nbh, "marking uptodate");
++info->nr_replays;
/* ll_rw_block(WRITE, 1, &nbh); */
unlock_buffer(nbh);
brelse(obh);
brelse(nbh);
}
skip_write:
tagp += tag_bytes;
if (!(flags & JFS_FLAG_SAME_UUID))
tagp += 16;
if (flags & JFS_FLAG_LAST_TAG)
break;
}
brelse(bh);
continue;
case JFS_COMMIT_BLOCK:
/* How to differentiate between interrupted commit
* and journal corruption ?
*
* {nth transaction}
* Checksum Verification Failed
* |
* ____________________
* | |
* async_commit sync_commit
* | |
* | GO TO NEXT "Journal Corruption"
* | TRANSACTION
* |
* {(n+1)th transanction}
* |
* _______|______________
* | |
* Commit block found Commit block not found
* | |
* "Journal Corruption" |
* _____________|_________
* | |
* nth trans corrupt OR nth trans
* and (n+1)th interrupted interrupted
* before commit block
* could reach the disk.
* (Cannot find the difference in above
* mentioned conditions. Hence assume
* "Interrupted Commit".)
*/
/* Found an expected commit block: if checksums
* are present verify them in PASS_SCAN; else not
* much to do other than move on to the next sequence
* number. */
if (pass == PASS_SCAN &&
jfs_has_feature_checksum(journal)) {
int chksum_err, chksum_seen;
struct commit_header *cbh =
(struct commit_header *)bh->b_data;
unsigned found_chksum =
ext2fs_be32_to_cpu(cbh->h_chksum[0]);
chksum_err = chksum_seen = 0;
if (info->end_transaction) {
journal->j_failed_commit =
info->end_transaction;
brelse(bh);
break;
}
if (crc32_sum == found_chksum &&
cbh->h_chksum_type == JFS_CRC32_CHKSUM &&
cbh->h_chksum_size ==
JFS_CRC32_CHKSUM_SIZE)
chksum_seen = 1;
else if (!(cbh->h_chksum_type == 0 &&
cbh->h_chksum_size == 0 &&
found_chksum == 0 &&
!chksum_seen))
/*
* If fs is mounted using an old kernel and then
* kernel with journal_chksum is used then we
* get a situation where the journal flag has
* checksum flag set but checksums are not
* present i.e chksum = 0, in the individual
* commit blocks.
* Hence to avoid checksum failures, in this
* situation, this extra check is added.
*/
chksum_err = 1;
if (chksum_err) {
info->end_transaction = next_commit_ID;
if (!jfs_has_feature_async_commit(journal)){
journal->j_failed_commit =
next_commit_ID;
brelse(bh);
break;
}
}
crc32_sum = ~0;
}
if (pass == PASS_SCAN &&
!jbd2_commit_block_csum_verify(journal,
bh->b_data)) {
info->end_transaction = next_commit_ID;
if (!jfs_has_feature_async_commit(journal)) {
journal->j_failed_commit =
next_commit_ID;
brelse(bh);
break;
}
}
brelse(bh);
next_commit_ID++;
continue;
case JFS_REVOKE_BLOCK:
/* If we aren't in the REVOKE pass, then we can
* just skip over this block. */
if (pass != PASS_REVOKE) {
brelse(bh);
continue;
}
err = scan_revoke_records(journal, bh,
next_commit_ID, info);
brelse(bh);
if (err)
goto failed;
continue;
default:
jbd_debug(3, "Unrecognised magic %d, end of scan.\n",
blocktype);
brelse(bh);
goto done;
}
}
done:
/*
* We broke out of the log scan loop: either we came to the
* known end of the log or we found an unexpected block in the
* log. If the latter happened, then we know that the "current"
* transaction marks the end of the valid log.
*/
if (pass == PASS_SCAN) {
if (!info->end_transaction)
info->end_transaction = next_commit_ID;
} else {
/* It's really bad news if different passes end up at
* different places (but possible due to IO errors). */
if (info->end_transaction != next_commit_ID) {
printk(KERN_ERR "JBD2: recovery pass %d ended at "
"transaction %u, expected %u\n",
pass, next_commit_ID, info->end_transaction);
if (!success)
success = -EIO;
}
}
if (block_error && success == 0)
success = -EIO;
return success;
failed:
return err;
}
errcode_t journal_find_head(journal_t *journal)
{
unsigned int next_commit_ID;
blk64_t next_log_block, head_block;
int err;
journal_superblock_t *sb;
journal_header_t *tmp;
struct buffer_head *bh;
unsigned int sequence;
int blocktype;
/*
* First thing is to establish what we expect to find in the log
* (in terms of transaction IDs), and where (in terms of log
* block offsets): query the superblock.
*/
sb = journal->j_superblock;
next_commit_ID = ext2fs_be32_to_cpu(sb->s_sequence);
next_log_block = ext2fs_be32_to_cpu(sb->s_start);
head_block = next_log_block;
if (next_log_block == 0)
return 0;
bh = getblk(journal->j_dev, 0, journal->j_blocksize);
if (bh == NULL)
return ENOMEM;
/*
* Now we walk through the log, transaction by transaction,
* making sure that each transaction has a commit block in the
* expected place. Each complete transaction gets replayed back
* into the main filesystem.
*/
while (1) {
dbg_printf("Scanning for sequence ID %u at %lu/%lu\n",
next_commit_ID, (unsigned long)next_log_block,
journal->j_last);
/* Skip over each chunk of the transaction looking
* either the next descriptor block or the final commit
* record. */
err = journal_bmap(journal, next_log_block, &bh->b_blocknr);
if (err)
goto err;
mark_buffer_uptodate(bh, 0);
ll_rw_block(READ, 1, &bh);
err = bh->b_err;
if (err)
goto err;
next_log_block++;
wrap(journal, next_log_block);
/* What kind of buffer is it?
*
* If it is a descriptor block, check that it has the
* expected sequence number. Otherwise, we're all done
* here. */
tmp = (journal_header_t *)bh->b_data;
if (tmp->h_magic != ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER)) {
dbg_printf("JBD2: wrong magic 0x%x\n", tmp->h_magic);
goto err;
}
blocktype = ext2fs_be32_to_cpu(tmp->h_blocktype);
sequence = ext2fs_be32_to_cpu(tmp->h_sequence);
dbg_printf("Found magic %d, sequence %d\n",
blocktype, sequence);
if (sequence != next_commit_ID) {
dbg_printf("JBD2: Wrong sequence %d (wanted %d)\n",
sequence, next_commit_ID);
goto err;
}
/* OK, we have a valid descriptor block which matches
* all of the sequence number checks. What are we going
* to do with it? That depends on the pass... */
switch (blocktype) {
case JFS_DESCRIPTOR_BLOCK:
next_log_block += count_tags(journal, bh->b_data);
wrap(journal, next_log_block);
continue;
case JFS_COMMIT_BLOCK:
head_block = next_log_block;
next_commit_ID++;
continue;
case JFS_REVOKE_BLOCK:
continue;
default:
dbg_printf("Unrecognised magic %d, end of scan.\n",
blocktype);
err = -EINVAL;
goto err;
}
}
err:
if (err == 0) {
dbg_printf("head seq=%d blk=%llu\n", next_commit_ID,
head_block);
journal->j_transaction_sequence = next_commit_ID;
journal->j_head = head_block;
}
brelse(bh);
return err;
}
