static void update_uuid(journal_t *journal)
{
size_t z;
ext2_filsys fs;
if (journal->j_format_version < 2)
return;
for (z = 0; z < sizeof(journal->j_superblock->s_uuid); z++)
if (journal->j_superblock->s_uuid[z])
break;
if (z == 0)
return;
fs = journal->j_fs_dev->k_fs;
if (!EXT2_HAS_INCOMPAT_FEATURE(fs->super,
EXT4_FEATURE_INCOMPAT_64BIT))
return;
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT) &&
EXT2_HAS_INCOMPAT_FEATURE(fs->super,
EXT4_FEATURE_INCOMPAT_64BIT))
return;
if (journal->j_tail != 0 ||
EXT2_HAS_INCOMPAT_FEATURE(fs->super,
EXT3_FEATURE_INCOMPAT_RECOVER)) {
printf("Journal needs recovery, will not set 64bit.\n");
return;
}
memcpy(journal->j_superblock->s_uuid, fs->super->s_uuid,
sizeof(fs->super->s_uuid));
}
static int count_tags(journal_t *journal, struct buffer_head *bh)
{
char * tagp;
journal_block_tag_t * tag;
int nr = 0, size = journal->j_blocksize;
int tag_bytes = journal_tag_bytes(journal);
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_CSUM_V2))
size -= sizeof(struct journal_block_tail);
tagp = &bh->b_data[sizeof(journal_header_t)];
while ((tagp - bh->b_data + tag_bytes) <= size) {
tag = (journal_block_tag_t *) tagp;
nr++;
tagp += tag_bytes;
if (!(tag->t_flags & cpu_to_be16(JFS_FLAG_SAME_UUID)))
tagp += 16;
if (tag->t_flags & cpu_to_be16(JFS_FLAG_LAST_TAG))
break;
}
return nr;
}
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 = be32_to_cpu(header->r_count);
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT))
record_len = 8;
while (offset < max) {
unsigned long long blocknr;
int err;
if (record_len == 4)
blocknr = ext2fs_be32_to_cpu(*((__be32 *)(bh->b_data +
offset)));
else
blocknr = ext2fs_be64_to_cpu(*((__be64 *)(bh->b_data +
offset)));
offset += record_len;
err = journal_set_revoke(journal, blocknr, sequence);
if (err)
return err;
++info->nr_revokes;
}
return 0;
}
static blk64_t journal_guess_blocks(journal_t *journal, blk64_t data_blocks,
blk64_t revoke_blocks)
{
blk64_t ret = 1;
unsigned int bs, sz;
/* Estimate # of revoke blocks */
bs = journal->j_blocksize;
if (journal_has_csum_v2or3(journal))
bs -= sizeof(struct journal_revoke_tail);
sz = JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT) ?
sizeof(__u64) : sizeof(__u32);
ret += revoke_blocks * sz / bs;
/* Estimate # of data blocks */
bs = journal->j_blocksize - 16;
if (journal_has_csum_v2or3(journal))
bs -= sizeof(struct journal_block_tail);
sz = journal_tag_bytes(journal);
ret += data_blocks * sz / bs;
ret += data_blocks;
return ret;
}
/* Checksumming functions */
static int e2fsck_journal_verify_csum_type(journal_t *j,
journal_superblock_t *jsb)
{
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 1;
return jsb->s_checksum_type == JBD2_CRC32C_CHKSUM;
}
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;
}
static errcode_t e2fsck_journal_sb_csum_set(journal_t *j,
journal_superblock_t *jsb)
{
__u32 crc;
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 0;
crc = e2fsck_journal_sb_csum(jsb);
jsb->s_checksum = ext2fs_cpu_to_be32(crc);
return 0;
}
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_block_tag_csum_verify(journal_t *j, journal_block_tag_t *tag,
void *buf, __u32 sequence)
{
__u32 provided, calculated;
if (!JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V2))
return 1;
sequence = ext2fs_cpu_to_be32(sequence);
calculated = ext2fs_crc32c_le(~0, j->j_superblock->s_uuid,
sizeof(j->j_superblock->s_uuid));
calculated = ext2fs_crc32c_le(calculated, (__u8 *)&sequence,
sizeof(sequence));
calculated = ext2fs_crc32c_le(calculated, buf, j->j_blocksize) & 0xffff;
provided = ext2fs_be16_to_cpu(tag->t_checksum);
return provided == ext2fs_cpu_to_be32(calculated);
}
static int jbd2_block_tag_csum_verify(journal_t *j, journal_block_tag_t *tag,
void *buf, __u32 sequence)
{
journal_block_tag3_t *tag3 = (journal_block_tag3_t *)tag;
__u32 csum32;
__u32 seq;
if (!journal_has_csum_v2or3(j))
return 1;
seq = ext2fs_cpu_to_be32(sequence);
csum32 = jbd2_chksum(j, j->j_csum_seed, (__u8 *)&seq, sizeof(seq));
csum32 = jbd2_chksum(j, csum32, buf, j->j_blocksize);
if (JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V3))
return tag3->t_checksum == ext2fs_cpu_to_be32(csum32);
return tag->t_checksum == ext2fs_cpu_to_be16(csum32);
}
void jbd2_block_tag_csum_set(journal_t *j, journal_block_tag_t *tag,
struct buffer_head *bh, __u32 sequence)
{
journal_block_tag3_t *tag3 = (journal_block_tag3_t *)tag;
__u32 csum32;
__be32 seq;
if (!journal_has_csum_v2or3(j))
return;
seq = ext2fs_cpu_to_be32(sequence);
csum32 = jbd2_chksum(j, j->j_csum_seed, (__u8 *)&seq, sizeof(seq));
csum32 = jbd2_chksum(j, csum32, bh->b_data, bh->b_size);
if (JFS_HAS_INCOMPAT_FEATURE(j, JFS_FEATURE_INCOMPAT_CSUM_V3))
tag3->t_checksum = ext2fs_cpu_to_be32(csum32);
else
tag->t_checksum = ext2fs_cpu_to_be16(csum32);
}
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;
}
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 update_64bit_flag(journal_t *journal)
{
if (journal->j_format_version < 2)
return;
if (!EXT2_HAS_INCOMPAT_FEATURE(journal->j_fs_dev->k_fs->super,
EXT4_FEATURE_INCOMPAT_64BIT))
return;
if (JFS_HAS_INCOMPAT_FEATURE(journal, JFS_FEATURE_INCOMPAT_64BIT) &&
EXT2_HAS_INCOMPAT_FEATURE(journal->j_fs_dev->k_fs->super,
EXT4_FEATURE_INCOMPAT_64BIT))
return;
if (journal->j_tail != 0 ||
EXT2_HAS_INCOMPAT_FEATURE(journal->j_fs_dev->k_fs->super,
EXT3_FEATURE_INCOMPAT_RECOVER)) {
printf("Journal needs recovery, will not set 64bit.\n");
return;
}
journal->j_superblock->s_feature_incompat |=
ext2fs_cpu_to_be32(JFS_FEATURE_INCOMPAT_64BIT);
}
static errcode_t e2fsck_journal_load(journal_t *journal)
{
e2fsck_t ctx = journal->j_dev->k_ctx;
journal_superblock_t *jsb;
struct buffer_head *jbh = journal->j_sb_buffer;
struct problem_context pctx;
clear_problem_context(&pctx);
ll_rw_block(READ, 1, &jbh);
if (jbh->b_err) {
com_err(ctx->device_name, jbh->b_err,
_("reading journal superblock\n"));
return jbh->b_err;
}
jsb = journal->j_superblock;
/* If we don't even have JFS_MAGIC, we probably have a wrong inode */
if (jsb->s_header.h_magic != htonl(JFS_MAGIC_NUMBER))
return e2fsck_journal_fix_bad_inode(ctx, &pctx);
switch (ntohl(jsb->s_header.h_blocktype)) {
case JFS_SUPERBLOCK_V1:
journal->j_format_version = 1;
if (jsb->s_feature_compat ||
jsb->s_feature_incompat ||
jsb->s_feature_ro_compat ||
jsb->s_nr_users)
clear_v2_journal_fields(journal);
break;
case JFS_SUPERBLOCK_V2:
journal->j_format_version = 2;
if (ntohl(jsb->s_nr_users) > 1 &&
uuid_is_null(ctx->fs->super->s_journal_uuid))
clear_v2_journal_fields(journal);
if (ntohl(jsb->s_nr_users) > 1) {
fix_problem(ctx, PR_0_JOURNAL_UNSUPP_MULTIFS, &pctx);
return EXT2_ET_JOURNAL_UNSUPP_VERSION;
}
break;
/*
* These should never appear in a journal super block, so if
* they do, the journal is badly corrupted.
*/
case JFS_DESCRIPTOR_BLOCK:
case JFS_COMMIT_BLOCK:
case JFS_REVOKE_BLOCK:
return EXT2_ET_CORRUPT_SUPERBLOCK;
/* If we don't understand the superblock major type, but there
* is a magic number, then it is likely to be a new format we
* just don't understand, so leave it alone. */
default:
return EXT2_ET_JOURNAL_UNSUPP_VERSION;
}
if (JFS_HAS_INCOMPAT_FEATURE(journal, ~JFS_KNOWN_INCOMPAT_FEATURES))
return EXT2_ET_UNSUPP_FEATURE;
if (JFS_HAS_RO_COMPAT_FEATURE(journal, ~JFS_KNOWN_ROCOMPAT_FEATURES))
return EXT2_ET_RO_UNSUPP_FEATURE;
/* We have now checked whether we know enough about the journal
* format to be able to proceed safely, so any other checks that
* fail we should attempt to recover from. */
if (jsb->s_blocksize != htonl(journal->j_blocksize)) {
com_err(ctx->program_name, EXT2_ET_CORRUPT_SUPERBLOCK,
_("%s: no valid journal superblock found\n"),
ctx->device_name);
return EXT2_ET_CORRUPT_SUPERBLOCK;
}
if (ntohl(jsb->s_maxlen) < journal->j_maxlen)
journal->j_maxlen = ntohl(jsb->s_maxlen);
else if (ntohl(jsb->s_maxlen) > journal->j_maxlen) {
com_err(ctx->program_name, EXT2_ET_CORRUPT_SUPERBLOCK,
_("%s: journal too short\n"),
ctx->device_name);
return EXT2_ET_CORRUPT_SUPERBLOCK;
}
journal->j_tail_sequence = ntohl(jsb->s_sequence);
journal->j_transaction_sequence = journal->j_tail_sequence;
journal->j_tail = ntohl(jsb->s_start);
journal->j_first = ntohl(jsb->s_first);
journal->j_last = ntohl(jsb->s_maxlen);
return 0;
}
static errcode_t journal_add_revoke_to_trans(journal_transaction_t *trans,
blk64_t *revoke_list,
size_t revoke_len)
{
journal_revoke_header_t *jrb;
void *buf;
size_t i, offset;
blk64_t curr_blk;
int sz, csum_size = 0;
struct buffer_head *bh;
errcode_t err;
JOURNAL_CHECK_TRANS_MAGIC(trans);
if ((trans->flags & J_TRANS_COMMITTED) ||
!(trans->flags & J_TRANS_OPEN))
return EXT2_ET_INVALID_ARGUMENT;
if (revoke_len == 0)
return 0;
/* Do we need to leave space at the end for a checksum? */
if (journal_has_csum_v2or3(trans->journal))
csum_size = sizeof(struct journal_revoke_tail);
curr_blk = trans->block;
bh = getblk(trans->journal->j_dev, curr_blk,
trans->journal->j_blocksize);
if (bh == NULL)
return ENOMEM;
jrb = buf = bh->b_data;
jrb->r_header.h_magic = ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER);
jrb->r_header.h_blocktype = ext2fs_cpu_to_be32(JFS_REVOKE_BLOCK);
jrb->r_header.h_sequence = ext2fs_cpu_to_be32(trans->tid);
offset = sizeof(*jrb);
if (JFS_HAS_INCOMPAT_FEATURE(trans->journal,
JFS_FEATURE_INCOMPAT_64BIT))
sz = 8;
else
sz = 4;
for (i = 0; i < revoke_len; i++) {
/* Block full, write to journal */
if (offset + sz > trans->journal->j_blocksize - csum_size) {
jrb->r_count = ext2fs_cpu_to_be32(offset);
jbd2_revoke_csum_set(trans->journal, bh);
err = journal_bmap(trans->journal, curr_blk,
&bh->b_blocknr);
if (err)
goto error;
dbg_printf("Writing revoke block at %llu:%llu\n",
curr_blk, bh->b_blocknr);
mark_buffer_dirty(bh);
ll_rw_block(WRITE, 1, &bh);
err = bh->b_err;
if (err)
goto error;
offset = sizeof(*jrb);
curr_blk++;
}
if (revoke_list[i] >=
ext2fs_blocks_count(trans->journal->j_fs_dev->k_fs->super)) {
err = EXT2_ET_BAD_BLOCK_NUM;
goto error;
}
if (JFS_HAS_INCOMPAT_FEATURE(trans->journal,
JFS_FEATURE_INCOMPAT_64BIT))
*((__u64 *)(&((char *)buf)[offset])) =
ext2fs_cpu_to_be64(revoke_list[i]);
else
*((__u32 *)(&((char *)buf)[offset])) =
ext2fs_cpu_to_be32(revoke_list[i]);
offset += sz;
}
if (offset > 0) {
jrb->r_count = ext2fs_cpu_to_be32(offset);
jbd2_revoke_csum_set(trans->journal, bh);
err = journal_bmap(trans->journal, curr_blk, &bh->b_blocknr);
if (err)
goto error;
dbg_printf("Writing revoke block at %llu:%llu\n",
curr_blk, bh->b_blocknr);
mark_buffer_dirty(bh);
ll_rw_block(WRITE, 1, &bh);
err = bh->b_err;
if (err)
goto error;
curr_blk++;
}
error:
trans->block = curr_blk;
brelse(bh);
return err;
}
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 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 */
/* Precompute the maximum metadata descriptors in a descriptor block */
int MAX_BLOCKS_PER_DESC;
MAX_BLOCKS_PER_DESC = ((journal->j_blocksize-sizeof(journal_header_t))
/ tag_bytes);
/*
* 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 = be32_to_cpu(sb->s_sequence);
next_log_block = 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, "JBD: 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 != cpu_to_be32(JFS_MAGIC_NUMBER)) {
brelse(bh);
break;
}
blocktype = be32_to_cpu(tmp->h_blocktype);
sequence = 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:
/* 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_COMPAT_FEATURE(journal,
JFS_FEATURE_COMPAT_CHECKSUM) &&
!info->end_transaction) {
if (calc_chksums(journal, bh,
&next_log_block,
&crc32_sum)) {
brelse(bh);
break;
}
brelse(bh);
continue;
}
next_log_block += count_tags(journal, bh);
wrap(journal, next_log_block);
brelse(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) {
unsigned long long io_block;
tag = (journal_block_tag_t *) tagp;
flags = be32_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
"JBD: IO error %d recovering "
"block %llu in log\n",
err, io_block);
} else {
unsigned long long blocknr;
J_ASSERT(obh != NULL);
blocknr = read_tag_block(tag_bytes,
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;
}
/* 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
"JBD: 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) {
journal_header_t *header;
header = (journal_header_t *) &nbh->b_data[0];
header->h_magic = 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:
jbd_debug(3, "Commit block for #%u found\n",
next_commit_ID);
/* 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_COMPAT_FEATURE(journal,
JFS_FEATURE_COMPAT_CHECKSUM)) {
int chksum_err, chksum_seen;
struct commit_header *cbh =
(struct commit_header *)bh->b_data;
unsigned found_chksum =
be32_to_cpu(cbh->h_chksum[0]);
chksum_err = chksum_seen = 0;
jbd_debug(3, "Checksums %x %x\n",
crc32_sum, found_chksum);
if (info->end_transaction) {
journal->j_failed_commit =
info->end_transaction;
brelse(bh);
break;
}
if (crc32_sum == found_chksum &&
cbh->h_chksum_type == JBD2_CRC32_CHKSUM &&
cbh->h_chksum_size ==
JBD2_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;
jbd_debug(1, "Checksum_err %x %x\n",
crc32_sum, found_chksum);
if (!JFS_HAS_INCOMPAT_FEATURE(journal,
JFS_FEATURE_INCOMPAT_ASYNC_COMMIT)){
journal->j_failed_commit =
next_commit_ID;
brelse(bh);
break;
}
}
crc32_sum = ~0;
}
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 "JBD: recovery pass %d ended at "
"transaction %u, expected %u\n",
pass, next_commit_ID, info->end_transaction);
if (!success)
success = -EIO;
}
}
return success;
failed:
return err;
}
static errcode_t journal_add_blocks_to_trans(journal_transaction_t *trans,
blk64_t *block_list, size_t block_len,
FILE *fp)
{
blk64_t curr_blk, jdb_blk;
size_t i, j;
int csum_size = 0;
journal_header_t *jdb;
journal_block_tag_t *jdbt;
int tag_bytes;
void *buf = NULL, *jdb_buf = NULL;
struct buffer_head *bh = NULL, *data_bh;
errcode_t err;
JOURNAL_CHECK_TRANS_MAGIC(trans);
if ((trans->flags & J_TRANS_COMMITTED) ||
!(trans->flags & J_TRANS_OPEN))
return EXT2_ET_INVALID_ARGUMENT;
if (block_len == 0)
return 0;
/* Do we need to leave space at the end for a checksum? */
if (journal_has_csum_v2or3(trans->journal))
csum_size = sizeof(struct journal_block_tail);
curr_blk = jdb_blk = trans->block;
data_bh = getblk(trans->journal->j_dev, curr_blk,
trans->journal->j_blocksize);
if (data_bh == NULL)
return ENOMEM;
buf = data_bh->b_data;
/* write the descriptor block header */
bh = getblk(trans->journal->j_dev, curr_blk,
trans->journal->j_blocksize);
if (bh == NULL) {
err = ENOMEM;
goto error;
}
jdb = jdb_buf = bh->b_data;
jdb->h_magic = ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER);
jdb->h_blocktype = ext2fs_cpu_to_be32(JFS_DESCRIPTOR_BLOCK);
jdb->h_sequence = ext2fs_cpu_to_be32(trans->tid);
jdbt = (journal_block_tag_t *)(jdb + 1);
curr_blk++;
for (i = 0; i < block_len; i++) {
j = fread(data_bh->b_data, trans->journal->j_blocksize, 1, fp);
if (j != 1) {
err = errno;
goto error;
}
tag_bytes = journal_tag_bytes(trans->journal);
/* No space left in descriptor block, write it out */
if ((char *)jdbt + tag_bytes >
(char *)jdb_buf + trans->journal->j_blocksize - csum_size) {
jbd2_descr_block_csum_set(trans->journal, bh);
err = journal_bmap(trans->journal, jdb_blk,
&bh->b_blocknr);
if (err)
goto error;
dbg_printf("Writing descriptor block at %llu:%llu\n",
jdb_blk, bh->b_blocknr);
mark_buffer_dirty(bh);
ll_rw_block(WRITE, 1, &bh);
err = bh->b_err;
if (err)
goto error;
jdbt = (journal_block_tag_t *)(jdb + 1);
jdb_blk = curr_blk;
curr_blk++;
}
if (block_list[i] >=
ext2fs_blocks_count(trans->journal->j_fs_dev->k_fs->super)) {
err = EXT2_ET_BAD_BLOCK_NUM;
goto error;
}
/* Fill out the block tag */
jdbt->t_blocknr = ext2fs_cpu_to_be32(block_list[i] & 0xFFFFFFFF);
jdbt->t_flags = 0;
if (jdbt != (journal_block_tag_t *)(jdb + 1))
jdbt->t_flags |= ext2fs_cpu_to_be16(JFS_FLAG_SAME_UUID);
else {
memcpy(jdbt + tag_bytes,
trans->journal->j_superblock->s_uuid,
sizeof(trans->journal->j_superblock->s_uuid));
tag_bytes += 16;
}
if (i == block_len - 1)
jdbt->t_flags |= ext2fs_cpu_to_be16(JFS_FLAG_LAST_TAG);
if (*((__u32 *)buf) == ext2fs_cpu_to_be32(JFS_MAGIC_NUMBER)) {
*((__u32 *)buf) = 0;
jdbt->t_flags |= ext2fs_cpu_to_be16(JFS_FLAG_ESCAPE);
}
if (JFS_HAS_INCOMPAT_FEATURE(trans->journal,
JFS_FEATURE_INCOMPAT_64BIT))
jdbt->t_blocknr_high = ext2fs_cpu_to_be32(block_list[i] >> 32);
jbd2_block_tag_csum_set(trans->journal, jdbt, data_bh,
trans->tid);
/* Write the data block */
err = journal_bmap(trans->journal, curr_blk,
&data_bh->b_blocknr);
if (err)
goto error;
dbg_printf("Writing data block %llu at %llu:%llu tag %d\n",
block_list[i], curr_blk, data_bh->b_blocknr,
tag_bytes);
mark_buffer_dirty(data_bh);
ll_rw_block(WRITE, 1, &data_bh);
err = data_bh->b_err;
if (err)
goto error;
curr_blk++;
jdbt = (journal_block_tag_t *)(((char *)jdbt) + tag_bytes);
}
/* Write out the last descriptor block */
if (jdbt != (journal_block_tag_t *)(jdb + 1)) {
jbd2_descr_block_csum_set(trans->journal, bh);
err = journal_bmap(trans->journal, jdb_blk, &bh->b_blocknr);
if (err)
goto error;
dbg_printf("Writing descriptor block at %llu:%llu\n",
jdb_blk, bh->b_blocknr);
mark_buffer_dirty(bh);
ll_rw_block(WRITE, 1, &bh);
err = bh->b_err;
if (err)
goto error;
}
error:
trans->block = curr_blk;
if (bh)
brelse(bh);
brelse(data_bh);
return err;
}
static errcode_t ext2fs_journal_load(journal_t *journal)
{
ext2_filsys fs = journal->j_dev->k_fs;
journal_superblock_t *jsb;
struct buffer_head *jbh = journal->j_sb_buffer;
ll_rw_block(READ, 1, &jbh);
if (jbh->b_err)
return jbh->b_err;
jsb = journal->j_superblock;
/* If we don't even have JFS_MAGIC, we probably have a wrong inode */
if (jsb->s_header.h_magic != htonl(JFS_MAGIC_NUMBER))
return ext2fs_journal_fix_bad_inode(fs);
switch (ntohl(jsb->s_header.h_blocktype)) {
case JFS_SUPERBLOCK_V1:
journal->j_format_version = 1;
if (jsb->s_feature_compat ||
jsb->s_feature_incompat ||
jsb->s_feature_ro_compat ||
jsb->s_nr_users)
clear_v2_journal_fields(journal);
break;
case JFS_SUPERBLOCK_V2:
journal->j_format_version = 2;
if (ntohl(jsb->s_nr_users) > 1 &&
uuid_is_null(fs->super->s_journal_uuid))
clear_v2_journal_fields(journal);
if (ntohl(jsb->s_nr_users) > 1)
return EXT2_ET_JOURNAL_UNSUPP_VERSION;
break;
/*
* These should never appear in a journal super block, so if
* they do, the journal is badly corrupted.
*/
case JFS_DESCRIPTOR_BLOCK:
case JFS_COMMIT_BLOCK:
case JFS_REVOKE_BLOCK:
return EXT2_ET_CORRUPT_JOURNAL_SB;
/* If we don't understand the superblock major type, but there
* is a magic number, then it is likely to be a new format we
* just don't understand, so leave it alone. */
default:
return EXT2_ET_JOURNAL_UNSUPP_VERSION;
}
if (JFS_HAS_INCOMPAT_FEATURE(journal, ~JFS_KNOWN_INCOMPAT_FEATURES))
return EXT2_ET_UNSUPP_FEATURE;
if (JFS_HAS_RO_COMPAT_FEATURE(journal, ~JFS_KNOWN_ROCOMPAT_FEATURES))
return EXT2_ET_RO_UNSUPP_FEATURE;
/* Checksum v1-3 are mutually exclusive features. */
if (jfs_has_feature_csum2(journal) && jfs_has_feature_csum3(journal))
return EXT2_ET_CORRUPT_JOURNAL_SB;
if (journal_has_csum_v2or3(journal) &&
jfs_has_feature_checksum(journal))
return EXT2_ET_CORRUPT_JOURNAL_SB;
if (!ext2fs_journal_verify_csum_type(journal, jsb) ||
!ext2fs_journal_sb_csum_verify(journal, jsb))
return EXT2_ET_CORRUPT_JOURNAL_SB;
if (journal_has_csum_v2or3(journal))
journal->j_csum_seed = jbd2_chksum(journal, ~0, jsb->s_uuid,
sizeof(jsb->s_uuid));
/* We have now checked whether we know enough about the journal
* format to be able to proceed safely, so any other checks that
* fail we should attempt to recover from. */
if (jsb->s_blocksize != htonl(journal->j_blocksize))
return EXT2_ET_CORRUPT_JOURNAL_SB;
if (ntohl(jsb->s_maxlen) < journal->j_maxlen)
journal->j_maxlen = ntohl(jsb->s_maxlen);
else if (ntohl(jsb->s_maxlen) > journal->j_maxlen)
return EXT2_ET_CORRUPT_JOURNAL_SB;
journal->j_tail_sequence = ntohl(jsb->s_sequence);
journal->j_transaction_sequence = journal->j_tail_sequence;
journal->j_tail = ntohl(jsb->s_start);
journal->j_first = ntohl(jsb->s_first);
journal->j_last = ntohl(jsb->s_maxlen);
return 0;
}