static int nilfs_palloc_get_block(struct inode *inode, unsigned long blkoff,
int create,
void (*init_block)(struct inode *,
struct buffer_head *,
void *),
struct buffer_head **bhp,
struct nilfs_bh_assoc *prev,
spinlock_t *lock)
{
int ret;
spin_lock(lock);
if (prev->bh && blkoff == prev->blkoff) {
get_bh(prev->bh);
*bhp = prev->bh;
spin_unlock(lock);
return 0;
}
spin_unlock(lock);
ret = nilfs_mdt_get_block(inode, blkoff, create, init_block, bhp);
if (!ret) {
spin_lock(lock);
/*
* The following code must be safe for change of the
* cache contents during the get block call.
*/
brelse(prev->bh);
get_bh(*bhp);
prev->bh = *bhp;
prev->blkoff = blkoff;
spin_unlock(lock);
}
return ret;
}
static int
nilfs_mdt_submit_block(struct inode *inode, unsigned long blkoff,
int mode, struct buffer_head **out_bh)
{
struct buffer_head *bh;
unsigned long blknum = 0;
int ret = -ENOMEM;
bh = nilfs_grab_buffer(inode, inode->i_mapping, blkoff, 0);
if (unlikely(!bh))
goto failed;
ret = -EEXIST; /* internal code */
if (buffer_uptodate(bh))
goto out;
if (mode == READA) {
if (!trylock_buffer(bh)) {
ret = -EBUSY;
goto failed_bh;
}
} else /* mode == READ */
lock_buffer(bh);
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
goto out;
}
if (!buffer_mapped(bh)) { /* unused buffer */
ret = nilfs_bmap_lookup(NILFS_I(inode)->i_bmap, blkoff,
&blknum);
if (unlikely(ret)) {
unlock_buffer(bh);
goto failed_bh;
}
bh->b_bdev = NILFS_MDT(inode)->mi_nilfs->ns_bdev;
bh->b_blocknr = blknum;
set_buffer_mapped(bh);
}
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(mode, bh);
ret = 0;
out:
get_bh(bh);
*out_bh = bh;
failed_bh:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
brelse(bh);
failed:
return ret;
}
static int __process_buffer(journal_t *journal, struct journal_head *jh,
int *batch_count, transaction_t *transaction)
{
struct buffer_head *bh = jh2bh(jh);
int ret = 0;
if (buffer_locked(bh)) {
get_bh(bh);
spin_unlock(&journal->j_list_lock);
wait_on_buffer(bh);
BUFFER_TRACE(bh, "brelse");
__brelse(bh);
ret = 1;
} else if (jh->b_transaction != NULL) {
transaction_t *t = jh->b_transaction;
tid_t tid = t->t_tid;
transaction->t_chp_stats.cs_forced_to_close++;
spin_unlock(&journal->j_list_lock);
if (unlikely(journal->j_flags & JBD2_UNMOUNT))
printk(KERN_ERR "JBD2: %s: "
"Waiting for Godot: block %llu\n",
journal->j_devname,
(unsigned long long) bh->b_blocknr);
jbd2_log_start_commit(journal, tid);
jbd2_log_wait_commit(journal, tid);
ret = 1;
} else if (!buffer_dirty(bh)) {
ret = 1;
if (unlikely(buffer_write_io_error(bh)))
ret = -EIO;
get_bh(bh);
BUFFER_TRACE(bh, "remove from checkpoint");
__jbd2_journal_remove_checkpoint(jh);
spin_unlock(&journal->j_list_lock);
__brelse(bh);
} else {
BUFFER_TRACE(bh, "queue");
get_bh(bh);
J_ASSERT_BH(bh, !buffer_jwrite(bh));
journal->j_chkpt_bhs[*batch_count] = bh;
__buffer_relink_io(jh);
transaction->t_chp_stats.cs_written++;
(*batch_count)++;
if (*batch_count == JBD2_NR_BATCH) {
spin_unlock(&journal->j_list_lock);
__flush_batch(journal, batch_count);
ret = 1;
}
}
return ret;
}
static int
nilfs_mdt_submit_block(struct inode *inode, unsigned long blkoff,
int mode, int mode_flags, struct buffer_head **out_bh)
{
struct buffer_head *bh;
__u64 blknum = 0;
int ret = -ENOMEM;
bh = nilfs_grab_buffer(inode, inode->i_mapping, blkoff, 0);
if (unlikely(!bh))
goto failed;
ret = -EEXIST; /* internal code */
if (buffer_uptodate(bh))
goto out;
if (mode_flags & REQ_RAHEAD) {
if (!trylock_buffer(bh)) {
ret = -EBUSY;
goto failed_bh;
}
} else /* mode == READ */
lock_buffer(bh);
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
goto out;
}
ret = nilfs_bmap_lookup(NILFS_I(inode)->i_bmap, blkoff, &blknum);
if (unlikely(ret)) {
unlock_buffer(bh);
goto failed_bh;
}
map_bh(bh, inode->i_sb, (sector_t)blknum);
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(mode, mode_flags, bh);
ret = 0;
trace_nilfs2_mdt_submit_block(inode, inode->i_ino, blkoff, mode);
out:
get_bh(bh);
*out_bh = bh;
failed_bh:
unlock_page(bh->b_page);
put_page(bh->b_page);
brelse(bh);
failed:
return ret;
}
int nilfs_load_inode_block(struct nilfs_sb_info *sbi, struct inode *inode,
struct buffer_head **pbh)
{
struct nilfs_inode_info *ii = NILFS_I(inode);
int err;
spin_lock(&sbi->s_inode_lock);
/* Caller of this function MUST lock s_inode_lock */
if (ii->i_bh == NULL) {
spin_unlock(&sbi->s_inode_lock);
err = nilfs_ifile_get_inode_block(sbi->s_ifile, inode->i_ino,
pbh);
if (unlikely(err))
return err;
spin_lock(&sbi->s_inode_lock);
if (ii->i_bh == NULL)
ii->i_bh = *pbh;
else {
brelse(*pbh);
*pbh = ii->i_bh;
}
} else
*pbh = ii->i_bh;
get_bh(*pbh);
spin_unlock(&sbi->s_inode_lock);
return 0;
}
int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh)
{
struct the_nilfs *nilfs = inode->i_sb->s_fs_info;
struct nilfs_inode_info *ii = NILFS_I(inode);
int err;
spin_lock(&nilfs->ns_inode_lock);
if (ii->i_bh == NULL) {
spin_unlock(&nilfs->ns_inode_lock);
err = nilfs_ifile_get_inode_block(ii->i_root->ifile,
inode->i_ino, pbh);
if (unlikely(err))
return err;
spin_lock(&nilfs->ns_inode_lock);
if (ii->i_bh == NULL)
ii->i_bh = *pbh;
else {
brelse(*pbh);
*pbh = ii->i_bh;
}
} else
*pbh = ii->i_bh;
get_bh(*pbh);
spin_unlock(&nilfs->ns_inode_lock);
return 0;
}
/*
* Write the MMP block using WRITE_SYNC to try to get the block on-disk
* faster.
*/
static int write_mmp_block(struct super_block *sb, struct buffer_head *bh)
{
struct mmp_struct *mmp = (struct mmp_struct *)(bh->b_data);
/*
* We protect against freezing so that we don't create dirty buffers
* on frozen filesystem.
*/
sb_start_write(sb);
ext4_mmp_csum_set(sb, mmp);
mark_buffer_dirty(bh);
lock_buffer(bh);
bh->b_end_io = end_buffer_write_sync;
get_bh(bh);
#ifdef FEATURE_STORAGE_META_LOG
if( bh && bh->b_bdev && bh->b_bdev->bd_disk)
set_metadata_rw_status(bh->b_bdev->bd_disk->first_minor, WAIT_WRITE_CNT);
#endif
submit_bh(WRITE_SYNC | REQ_META | REQ_PRIO, bh);
wait_on_buffer(bh);
sb_end_write(sb);
if (unlikely(!buffer_uptodate(bh)))
return 1;
return 0;
}
static int read_mmp_block(struct super_block *sb, struct buffer_head **bh,
ext4_fsblk_t mmp_block)
{
struct mmp_struct *mmp;
if (*bh)
clear_buffer_uptodate(*bh);
if (!*bh)
*bh = sb_getblk(sb, mmp_block);
if (!*bh)
return -ENOMEM;
if (*bh) {
get_bh(*bh);
lock_buffer(*bh);
(*bh)->b_end_io = end_buffer_read_sync;
submit_bh(READ_SYNC, *bh);
wait_on_buffer(*bh);
if (!buffer_uptodate(*bh)) {
brelse(*bh);
*bh = NULL;
}
}
if (!*bh) {
ext4_warning(sb, "Error while reading MMP block %llu",
mmp_block);
return -EIO;
}
mmp = (struct mmp_struct *)((*bh)->b_data);
if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC)
return -EINVAL;
return 0;
}
int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
struct buffer_head **bhp)
{
struct gfs2_sbd *sdp = gl->gl_sbd;
struct buffer_head *bh;
if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
return -EIO;
*bhp = bh = gfs2_getbuf(gl, blkno, CREATE);
lock_buffer(bh);
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
return 0;
}
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(READ_SYNC | (1 << BIO_RW_META), bh);
if (!(flags & DIO_WAIT))
return 0;
wait_on_buffer(bh);
if (unlikely(!buffer_uptodate(bh))) {
struct gfs2_trans *tr = current->journal_info;
if (tr && tr->tr_touched)
gfs2_io_error_bh(sdp, bh);
brelse(bh);
return -EIO;
}
return 0;
}
int reiserfs_remove_entry (reiserfs_filsys_t fs, struct key * key)
{
struct path path;
struct tree_balance tb;
struct item_head * ih;
struct reiserfs_de_head * deh;
if (_search_by_entry_key (fs, key, &path) != POSITION_FOUND) {
pathrelse (&path);
return 1;
}
ih = get_ih (&path);
if (ih_entry_count (ih) == 1) {
_init_tb_struct (&tb, fs, &path, -(IH_SIZE + ih_item_len (ih)));
if (fix_nodes (M_DELETE, &tb, 0) != CARRY_ON) {
unfix_nodes (&tb);
return 1;
}
do_balance (&tb, 0, 0, M_DELETE, 0);
return 0;
}
deh = B_I_DEH (get_bh (&path), ih) + path.pos_in_item;
_init_tb_struct (&tb, fs, &path, -(DEH_SIZE + entry_length (ih, deh, path.pos_in_item)));
if (fix_nodes (M_CUT, &tb, 0) != CARRY_ON) {
unfix_nodes (&tb);
return 1;
}
do_balance (&tb, 0, 0, M_CUT, 0);
return 0;
}
static int nilfs_mdt_create_block(struct inode *inode, unsigned long block,
struct buffer_head **out_bh,
void (*init_block)(struct inode *,
struct buffer_head *,
void *))
{
struct the_nilfs *nilfs = NILFS_MDT(inode)->mi_nilfs;
struct super_block *sb = inode->i_sb;
struct nilfs_transaction_info ti;
struct buffer_head *bh;
int err;
if (!sb) {
/*
* Make sure this function is not called from any
* read-only context.
*/
if (!nilfs->ns_writer) {
WARN_ON(1);
err = -EROFS;
goto out;
}
sb = nilfs->ns_writer->s_super;
}
nilfs_transaction_begin(sb, &ti, 0);
err = -ENOMEM;
bh = nilfs_grab_buffer(inode, inode->i_mapping, block, 0);
if (unlikely(!bh))
goto failed_unlock;
err = -EEXIST;
if (buffer_uptodate(bh))
goto failed_bh;
wait_on_buffer(bh);
if (buffer_uptodate(bh))
goto failed_bh;
bh->b_bdev = nilfs->ns_bdev;
err = nilfs_mdt_insert_new_block(inode, block, bh, init_block);
if (likely(!err)) {
get_bh(bh);
*out_bh = bh;
}
failed_bh:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
brelse(bh);
failed_unlock:
if (likely(!err))
err = nilfs_transaction_commit(sb);
else
nilfs_transaction_abort(sb);
out:
return err;
}
/*
* We were unable to perform jbd_trylock_bh_state() inside j_list_lock.
* The caller must restart a list walk. Wait for someone else to run
* jbd_unlock_bh_state().
*/
static void jbd_sync_bh(journal_t *journal, struct buffer_head *bh)
{
get_bh(bh);
spin_unlock(&journal->j_list_lock);
jbd_lock_bh_state(bh);
jbd_unlock_bh_state(bh);
put_bh(bh);
}
/*
* Try to flush one buffer from the checkpoint list to disk.
*
* Return 1 if something happened which requires us to abort the current
* scan of the checkpoint list.
*
* Called with j_list_lock held and drops it if 1 is returned
* Called under jbd_lock_bh_state(jh2bh(jh)), and drops it
*/
static int __process_buffer(journal_t *journal, struct journal_head *jh,
struct buffer_head **bhs, int *batch_count)
{
struct buffer_head *bh = jh2bh(jh);
int ret = 0;
if (buffer_locked(bh)) {
atomic_inc(&bh->b_count);
spin_unlock(&journal->j_list_lock);
jbd_unlock_bh_state(bh);
wait_on_buffer(bh);
/* the journal_head may have gone by now */
BUFFER_TRACE(bh, "brelse");
__brelse(bh);
ret = 1;
} else if (jh->b_transaction != NULL) {
transaction_t *t = jh->b_transaction;
tid_t tid = t->t_tid;
spin_unlock(&journal->j_list_lock);
jbd_unlock_bh_state(bh);
log_start_commit(journal, tid);
log_wait_commit(journal, tid);
ret = 1;
} else if (!buffer_dirty(bh)) {
J_ASSERT_JH(jh, !buffer_jbddirty(bh));
BUFFER_TRACE(bh, "remove from checkpoint");
__journal_remove_checkpoint(jh);
spin_unlock(&journal->j_list_lock);
jbd_unlock_bh_state(bh);
journal_remove_journal_head(bh);
__brelse(bh);
ret = 1;
} else {
/*
* Important: we are about to write the buffer, and
* possibly block, while still holding the journal lock.
* We cannot afford to let the transaction logic start
* messing around with this buffer before we write it to
* disk, as that would break recoverability.
*/
BUFFER_TRACE(bh, "queue");
get_bh(bh);
J_ASSERT_BH(bh, !buffer_jwrite(bh));
set_buffer_jwrite(bh);
bhs[*batch_count] = bh;
__buffer_relink_io(jh);
jbd_unlock_bh_state(bh);
(*batch_count)++;
if (*batch_count == NR_BATCH) {
spin_unlock(&journal->j_list_lock);
__flush_batch(journal, bhs, batch_count);
ret = 1;
}
}
return ret;
}
static int ufs_fmp_run(struct device *dev, uint32_t mode, uint8_t *data,
uint32_t len, uint32_t write)
{
int ret = 0;
struct ufs_hba *hba;
struct ufs_fmp_work *work;
struct Scsi_Host *host;
static struct buffer_head *bh;
work = dev_get_drvdata(dev);
if (!work) {
dev_err(dev, "Fail to get work from platform device\n");
return -ENODEV;
}
host = work->host;
hba = shost_priv(host);
hba->self_test_mode = mode;
bh = __getblk(work->bdev, work->sector, FMP_BLK_SIZE);
if (!bh) {
dev_err(dev, "Fail to get block from bdev\n");
return -ENODEV;
}
hba->self_test_bh = bh;
get_bh(bh);
if (write == WRITE_MODE) {
memcpy(bh->b_data, data, len);
set_buffer_dirty(bh);
sync_dirty_buffer(bh);
if (buffer_req(bh) && !buffer_uptodate(bh)) {
dev_err(dev, "IO error syncing for FMP fips write\n");
ret = -EIO;
goto out;
}
memset(bh->b_data, 0, FMP_BLK_SIZE);
} else {
lock_buffer(bh);
bh->b_end_io = end_buffer_read_sync;
submit_bh(READ_SYNC, bh);
wait_on_buffer(bh);
if (unlikely(!buffer_uptodate(bh))) {
ret = -EIO;
goto out;
}
memcpy(data, bh->b_data, len);
}
out:
hba->self_test_mode = 0;
hba->self_test_bh = NULL;
put_bh(bh);
return ret;
}
int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
sector_t pblocknr, struct buffer_head **pbh)
{
struct buffer_head *bh;
struct inode *inode = NILFS_BTNC_I(btnc);
int err;
bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node);
if (unlikely(!bh))
return -ENOMEM;
err = -EEXIST; /* internal code */
if (buffer_uptodate(bh) || buffer_dirty(bh))
goto found;
if (pblocknr == 0) {
pblocknr = blocknr;
if (inode->i_ino != NILFS_DAT_INO) {
struct inode *dat =
nilfs_dat_inode(NILFS_I_NILFS(inode));
/* blocknr is a virtual block number */
err = nilfs_dat_translate(dat, blocknr, &pblocknr);
if (unlikely(err)) {
brelse(bh);
goto out_locked;
}
}
}
lock_buffer(bh);
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
err = -EEXIST; /* internal code */
goto found;
}
set_buffer_mapped(bh);
bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev;
bh->b_blocknr = pblocknr; /* set block address for read */
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(READ, bh);
bh->b_blocknr = blocknr; /* set back to the given block address */
err = 0;
found:
*pbh = bh;
out_locked:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
return err;
}
/*
* Helper function to set i_mode in memory and disk. Some call paths
* will not have di_bh or a journal handle to pass, in which case it
* will create it's own.
*/
static int ocfs2_acl_set_mode(struct inode *inode, struct buffer_head *di_bh,
handle_t *handle, umode_t new_mode)
{
int ret, commit_handle = 0;
struct ocfs2_dinode *di;
if (di_bh == NULL) {
ret = ocfs2_read_inode_block(inode, &di_bh);
if (ret) {
mlog_errno(ret);
goto out;
}
} else
get_bh(di_bh);
if (handle == NULL) {
handle = ocfs2_start_trans(OCFS2_SB(inode->i_sb),
OCFS2_INODE_UPDATE_CREDITS);
if (IS_ERR(handle)) {
ret = PTR_ERR(handle);
mlog_errno(ret);
goto out_brelse;
}
commit_handle = 1;
}
di = (struct ocfs2_dinode *)di_bh->b_data;
ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (ret) {
mlog_errno(ret);
goto out_commit;
}
inode->i_mode = new_mode;
inode->i_ctime = current_time(inode);
di->i_mode = cpu_to_le16(inode->i_mode);
di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec);
di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec);
ocfs2_update_inode_fsync_trans(handle, inode, 0);
ocfs2_journal_dirty(handle, di_bh);
out_commit:
if (commit_handle)
ocfs2_commit_trans(OCFS2_SB(inode->i_sb), handle);
out_brelse:
brelse(di_bh);
out:
return ret;
}
void jbd2_journal_refile_buffer(journal_t *journal, struct journal_head *jh)
{
struct buffer_head *bh = jh2bh(jh);
get_bh(bh);
jbd_lock_bh_state(bh);
spin_lock(&journal->j_list_lock);
__jbd2_journal_refile_buffer(jh);
jbd_unlock_bh_state(bh);
spin_unlock(&journal->j_list_lock);
__brelse(bh);
}
/*
* Write the MMP block using WRITE_SYNC to try to get the block on-disk
* faster.
*/
static int write_mmp_block(struct buffer_head *bh)
{
mark_buffer_dirty(bh);
lock_buffer(bh);
bh->b_end_io = end_buffer_write_sync;
get_bh(bh);
submit_bh(WRITE_SYNC, bh);
wait_on_buffer(bh);
if (unlikely(!buffer_uptodate(bh)))
return 1;
return 0;
}
/*
* nilfs_gccache_submit_read_data() - add data buffer and submit read request
* @inode - gc inode
* @blkoff - dummy offset treated as the key for the page cache
* @pbn - physical block number of the block
* @vbn - virtual block number of the block, 0 for non-virtual block
* @out_bh - indirect pointer to a buffer_head struct to receive the results
*
* Description: nilfs_gccache_submit_read_data() registers the data buffer
* specified by @pbn to the GC pagecache with the key @blkoff.
* This function sets @vbn (@pbn if @vbn is zero) in b_blocknr of the buffer.
*
* Return Value: On success, 0 is returned. On Error, one of the following
* negative error code is returned.
*
* %-EIO - I/O error.
*
* %-ENOMEM - Insufficient amount of memory available.
*
* %-ENOENT - The block specified with @pbn does not exist.
*/
int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff,
sector_t pbn, __u64 vbn,
struct buffer_head **out_bh)
{
struct buffer_head *bh;
int err;
bh = nilfs_grab_buffer(inode, inode->i_mapping, blkoff, 0);
if (unlikely(!bh))
return -ENOMEM;
if (buffer_uptodate(bh))
goto out;
if (pbn == 0) {
struct the_nilfs *nilfs = inode->i_sb->s_fs_info;
err = nilfs_dat_translate(nilfs->ns_dat, vbn, &pbn);
if (unlikely(err)) { /* -EIO, -ENOMEM, -ENOENT */
brelse(bh);
goto failed;
}
}
lock_buffer(bh);
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
goto out;
}
if (!buffer_mapped(bh)) {
bh->b_bdev = inode->i_sb->s_bdev;
set_buffer_mapped(bh);
}
bh->b_blocknr = pbn;
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(READ, bh);
if (vbn)
bh->b_blocknr = vbn;
out:
err = 0;
*out_bh = bh;
failed:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
return err;
}
static int __try_to_free_cp_buf(struct journal_head *jh)
{
int ret = 0;
struct buffer_head *bh = jh2bh(jh);
if (jh->b_transaction == NULL && !buffer_locked(bh) &&
!buffer_dirty(bh) && !buffer_write_io_error(bh)) {
get_bh(bh);
JBUFFER_TRACE(jh, "remove from checkpoint list");
ret = __jbd2_journal_remove_checkpoint(jh) + 1;
BUFFER_TRACE(bh, "release");
__brelse(bh);
}
return ret;
}
int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
struct inode *inode)
{
int ret = 0;
mlog_entry("(bh->b_blocknr = %llu, inode=%p)\n",
(unsigned long long)bh->b_blocknr, inode);
BUG_ON(bh->b_blocknr < OCFS2_SUPER_BLOCK_BLKNO);
BUG_ON(buffer_jbd(bh));
/* No need to check for a soft readonly file system here. non
* journalled writes are only ever done on system files which
* can get modified during recovery even if read-only. */
if (ocfs2_is_hard_readonly(osb)) {
ret = -EROFS;
goto out;
}
mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
lock_buffer(bh);
set_buffer_uptodate(bh);
/* remove from dirty list before I/O. */
clear_buffer_dirty(bh);
get_bh(bh); /* for end_buffer_write_sync() */
bh->b_end_io = end_buffer_write_sync;
submit_bh(WRITE, bh);
wait_on_buffer(bh);
if (buffer_uptodate(bh)) {
ocfs2_set_buffer_uptodate(inode, bh);
} else {
/* We don't need to remove the clustered uptodate
* information for this bh as it's not marked locally
* uptodate. */
ret = -EIO;
put_bh(bh);
}
mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
out:
mlog_exit(ret);
return ret;
}
static int ufs_fmp_run(struct device *dev, uint32_t mode, uint8_t *data,
uint32_t len, uint32_t write)
{
struct ufs_hba *hba;
struct ufs_fmp_work *work;
struct Scsi_Host *host;
static struct buffer_head *bh;
work = dev_get_drvdata(dev);
if (!work) {
dev_err(dev, "Fail to get work from platform device\n");
return -ENODEV;
}
host = work->host;
hba = shost_priv(host);
hba->self_test_mode = mode;
bh = __getblk(work->bdev, work->sector, FMP_BLK_SIZE);
if (!bh) {
dev_err(dev, "Fail to get block from bdev\n");
return -ENODEV;
}
hba->self_test_bh = bh;
get_bh(bh);
if (write == WRITE_MODE) {
memcpy(bh->b_data, data, len);
bh->b_state &= ~(1 << BH_Uptodate);
bh->b_state &= ~(1 << BH_Lock);
ll_rw_block(WRITE_FLUSH_FUA, 1, &bh);
wait_on_buffer(bh);
memset(bh->b_data, 0, FMP_BLK_SIZE);
} else {
bh->b_state &= ~(1 << BH_Uptodate);
bh->b_state &= ~(1 << BH_Lock);
ll_rw_block(READ_SYNC, 1, &bh);
wait_on_buffer(bh);
memcpy(data, bh->b_data, len);
}
put_bh(bh);
hba->self_test_mode = 0;
hba->self_test_bh = NULL;
return 0;
}
int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
struct ocfs2_caching_info *ci)
{
int ret = 0;
trace_ocfs2_write_block((unsigned long long)bh->b_blocknr, ci);
BUG_ON(bh->b_blocknr < OCFS2_SUPER_BLOCK_BLKNO);
BUG_ON(buffer_jbd(bh));
/* No need to check for a soft readonly file system here. non
* journalled writes are only ever done on system files which
* can get modified during recovery even if read-only. */
if (ocfs2_is_hard_readonly(osb)) {
ret = -EROFS;
mlog_errno(ret);
goto out;
}
ocfs2_metadata_cache_io_lock(ci);
lock_buffer(bh);
set_buffer_uptodate(bh);
/* remove from dirty list before I/O. */
clear_buffer_dirty(bh);
get_bh(bh); /* for end_buffer_write_sync() */
bh->b_end_io = end_buffer_write_sync;
submit_bh(REQ_OP_WRITE, 0, bh);
wait_on_buffer(bh);
if (buffer_uptodate(bh)) {
ocfs2_set_buffer_uptodate(ci, bh);
} else {
/* We don't need to remove the clustered uptodate
* information for this bh as it's not marked locally
* uptodate. */
ret = -EIO;
mlog_errno(ret);
}
ocfs2_metadata_cache_io_unlock(ci);
out:
return ret;
}
static int nilfs_mdt_create_block(struct inode *inode, unsigned long block,
struct buffer_head **out_bh,
void (*init_block)(struct inode *,
struct buffer_head *,
void *))
{
struct super_block *sb = inode->i_sb;
struct nilfs_transaction_info ti;
struct buffer_head *bh;
int err;
nilfs_transaction_begin(sb, &ti, 0);
err = -ENOMEM;
bh = nilfs_grab_buffer(inode, inode->i_mapping, block, 0);
if (unlikely(!bh))
goto failed_unlock;
err = -EEXIST;
if (buffer_uptodate(bh))
goto failed_bh;
wait_on_buffer(bh);
if (buffer_uptodate(bh))
goto failed_bh;
bh->b_bdev = sb->s_bdev;
err = nilfs_mdt_insert_new_block(inode, block, bh, init_block);
if (likely(!err)) {
get_bh(bh);
*out_bh = bh;
}
failed_bh:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
brelse(bh);
failed_unlock:
if (likely(!err))
err = nilfs_transaction_commit(sb);
else
nilfs_transaction_abort(sb);
return err;
}
/*
* Read the MMP block. It _must_ be read from disk and hence we clear the
* uptodate flag on the buffer.
*/
static int read_mmp_block(struct super_block *sb, struct buffer_head **bh,
ext4_fsblk_t mmp_block)
{
struct mmp_struct *mmp;
int ret;
if (*bh)
clear_buffer_uptodate(*bh);
/* This would be sb_bread(sb, mmp_block), except we need to be sure
* that the MD RAID device cache has been bypassed, and that the read
* is not blocked in the elevator. */
if (!*bh) {
*bh = sb_getblk(sb, mmp_block);
if (!*bh) {
ret = -ENOMEM;
goto warn_exit;
}
}
get_bh(*bh);
lock_buffer(*bh);
(*bh)->b_end_io = end_buffer_read_sync;
submit_bh(READ_SYNC | REQ_META | REQ_PRIO, *bh);
wait_on_buffer(*bh);
if (!buffer_uptodate(*bh)) {
ret = -EIO;
goto warn_exit;
}
mmp = (struct mmp_struct *)((*bh)->b_data);
if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC) {
ret = -EFSCORRUPTED;
goto warn_exit;
}
if (!ext4_mmp_csum_verify(sb, mmp)) {
ret = -EFSBADCRC;
goto warn_exit;
}
return 0;
warn_exit:
brelse(*bh);
*bh = NULL;
ext4_warning(sb, "Error %d while reading MMP block %llu",
ret, mmp_block);
return ret;
}
/*
* Write the MMP block using WRITE_SYNC to try to get the block on-disk
* faster.
*/
static int write_mmp_block(struct buffer_head *bh)
{
mark_buffer_dirty(bh);
lock_buffer(bh);
bh->b_end_io = end_buffer_write_sync;
get_bh(bh);
#ifdef FEATURE_STORAGE_META_LOG
set_metadata_rw_status(bh->b_bdev->bd_disk->first_minor, WAIT_WRITE_CNT);
#endif
submit_bh(WRITE_SYNC, bh);
wait_on_buffer(bh);
if (unlikely(!buffer_uptodate(bh)))
return 1;
return 0;
}
static int __try_to_free_cp_buf(struct journal_head *jh)
{
int ret = 0;
struct buffer_head *bh = jh2bh(jh);
if (jh->b_jlist == BJ_None && !buffer_locked(bh) &&
!buffer_dirty(bh) && !buffer_write_io_error(bh)) {
get_bh(bh);
JBUFFER_TRACE(jh, "remove from checkpoint list");
ret = __journal_remove_checkpoint(jh) + 1;
jbd_unlock_bh_state(bh);
BUFFER_TRACE(bh, "release");
__brelse(bh);
} else {
jbd_unlock_bh_state(bh);
}
return ret;
}
struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create)
{
struct address_space *mapping = gl->gl_aspace->i_mapping;
struct gfs2_sbd *sdp = gl->gl_sbd;
struct page *page;
struct buffer_head *bh;
unsigned int shift;
unsigned long index;
unsigned int bufnum;
shift = PAGE_CACHE_SHIFT - sdp->sd_sb.sb_bsize_shift;
index = blkno >> shift; /* convert block to page */
bufnum = blkno - (index << shift); /* block buf index within page */
if (create) {
for (;;) {
page = grab_cache_page(mapping, index);
if (page)
break;
yield();
}
} else {
page = find_lock_page(mapping, index);
if (!page)
return NULL;
}
if (!page_has_buffers(page))
create_empty_buffers(page, sdp->sd_sb.sb_bsize, 0);
/* Locate header for our buffer within our page */
for (bh = page_buffers(page); bufnum--; bh = bh->b_this_page)
/* Do nothing */;
get_bh(bh);
if (!buffer_mapped(bh))
map_bh(bh, sdp->sd_vfs, blkno);
unlock_page(page);
mark_page_accessed(page);
page_cache_release(page);
return bh;
}
/*
* Try to flush one buffer from the checkpoint list to disk.
*
* Return 1 if something happened which requires us to abort the current
* scan of the checkpoint list.
*
* Called with j_list_lock held.
* Called under jbd_lock_bh_state(jh2bh(jh)), and drops it
*/
static int __flush_buffer(journal_t *journal, struct journal_head *jh,
struct buffer_head **bhs, int *batch_count,
int *drop_count)
{
struct buffer_head *bh = jh2bh(jh);
int ret = 0;
if (buffer_dirty(bh) && !buffer_locked(bh) && jh->b_jlist == BJ_None) {
J_ASSERT_JH(jh, jh->b_transaction == NULL);
/*
* Important: we are about to write the buffer, and
* possibly block, while still holding the journal lock.
* We cannot afford to let the transaction logic start
* messing around with this buffer before we write it to
* disk, as that would break recoverability.
*/
BUFFER_TRACE(bh, "queue");
get_bh(bh);
J_ASSERT_BH(bh, !buffer_jwrite(bh));
set_buffer_jwrite(bh);
bhs[*batch_count] = bh;
jbd_unlock_bh_state(bh);
(*batch_count)++;
if (*batch_count == NR_BATCH) {
__flush_batch(journal, bhs, batch_count);
ret = 1;
}
} else {
int last_buffer = 0;
if (jh->b_cpnext == jh) {
/* We may be about to drop the transaction. Tell the
* caller that the lists have changed.
*/
last_buffer = 1;
}
if (__try_to_free_cp_buf(jh)) {
(*drop_count)++;
ret = last_buffer;
}
}
return ret;
}
/*
* Read the MMP block. It _must_ be read from disk and hence we clear the
* uptodate flag on the buffer.
*/
static int read_mmp_block(struct super_block *sb, struct buffer_head **bh,
ext4_fsblk_t mmp_block)
{
struct mmp_struct *mmp;
if (*bh)
clear_buffer_uptodate(*bh);
/* This would be sb_bread(sb, mmp_block), except we need to be sure
* that the MD RAID device cache has been bypassed, and that the read
* is not blocked in the elevator. */
if (!*bh)
*bh = sb_getblk(sb, mmp_block);
if (!*bh)
return -ENOMEM;
if (*bh) {
get_bh(*bh);
lock_buffer(*bh);
(*bh)->b_end_io = end_buffer_read_sync;
#ifdef FEATURE_STORAGE_META_LOG
if( (*bh) && (*bh)->b_bdev && (*bh)->b_bdev->bd_disk)
set_metadata_rw_status((*bh)->b_bdev->bd_disk->first_minor, WAIT_READ_CNT);
#endif
submit_bh(READ_SYNC | REQ_META | REQ_PRIO, *bh);
wait_on_buffer(*bh);
if (!buffer_uptodate(*bh)) {
brelse(*bh);
*bh = NULL;
}
}
if (unlikely(!*bh)) {
ext4_warning(sb, "Error while reading MMP block %llu",
mmp_block);
return -EIO;
}
mmp = (struct mmp_struct *)((*bh)->b_data);
if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC ||
!ext4_mmp_csum_verify(sb, mmp))
return -EINVAL;
return 0;
}
