struct buffer_head *
nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr)
{
struct inode *inode = NILFS_BTNC_I(btnc);
struct buffer_head *bh;
bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node);
if (unlikely(!bh))
return NULL;
if (unlikely(buffer_mapped(bh) || buffer_uptodate(bh) ||
buffer_dirty(bh))) {
brelse(bh);
BUG();
}
memset(bh->b_data, 0, 1 << inode->i_blkbits);
bh->b_bdev = inode->i_sb->s_bdev;
bh->b_blocknr = blocknr;
set_buffer_mapped(bh);
set_buffer_uptodate(bh);
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
return bh;
}
/*
* write out a page to a file
*/
static void write_page(struct bitmap *bitmap, struct page *page, int wait)
{
struct buffer_head *bh;
if (bitmap->file == NULL) {
switch (write_sb_page(bitmap, page, wait)) {
case -EINVAL:
bitmap->flags |= BITMAP_WRITE_ERROR;
}
} else {
bh = page_buffers(page);
while (bh && bh->b_blocknr) {
atomic_inc(&bitmap->pending_writes);
set_buffer_locked(bh);
set_buffer_mapped(bh);
submit_bh(WRITE, bh);
bh = bh->b_this_page;
}
if (wait) {
wait_event(bitmap->write_wait,
atomic_read(&bitmap->pending_writes)==0);
}
}
if (bitmap->flags & BITMAP_WRITE_ERROR)
bitmap_file_kick(bitmap);
}
static int
nilfs_mdt_insert_new_block(struct inode *inode, unsigned long block,
struct buffer_head *bh,
void (*init_block)(struct inode *,
struct buffer_head *, void *))
{
struct nilfs_inode_info *ii = NILFS_I(inode);
void *kaddr;
int ret;
/* Caller exclude read accesses using page lock */
/* set_buffer_new(bh); */
bh->b_blocknr = 0;
ret = nilfs_bmap_insert(ii->i_bmap, block, (unsigned long)bh);
if (unlikely(ret))
return ret;
set_buffer_mapped(bh);
kaddr = kmap_atomic(bh->b_page, KM_USER0);
memset(kaddr + bh_offset(bh), 0, 1 << inode->i_blkbits);
if (init_block)
init_block(inode, bh, kaddr);
flush_dcache_page(bh->b_page);
kunmap_atomic(kaddr, KM_USER0);
set_buffer_uptodate(bh);
nilfs_mark_buffer_dirty(bh);
nilfs_mdt_mark_dirty(inode);
return 0;
}
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;
}
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;
}
/*
* 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;
}
/*
* write out a page to a file
*/
static int write_page(struct bitmap *bitmap, struct page *page, int wait)
{
struct buffer_head *bh;
if (bitmap->file == NULL)
return write_sb_page(bitmap, page, wait);
bh = page_buffers(page);
while (bh && bh->b_blocknr) {
atomic_inc(&bitmap->pending_writes);
set_buffer_locked(bh);
set_buffer_mapped(bh);
submit_bh(WRITE, bh);
bh = bh->b_this_page;
}
if (wait) {
wait_event(bitmap->write_wait,
atomic_read(&bitmap->pending_writes)==0);
return (bitmap->flags & BITMAP_WRITE_ERROR) ? -EIO : 0;
}
return 0;
}
int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
sector_t pblocknr, int mode,
struct buffer_head **pbh, sector_t *submit_ptr)
{
struct buffer_head *bh;
struct inode *inode = NILFS_BTNC_I(btnc);
struct page *page;
int err;
bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node);
if (unlikely(!bh))
return -ENOMEM;
err = -EEXIST;
page = bh->b_page;
if (buffer_uptodate(bh) || buffer_dirty(bh))
goto found;
if (pblocknr == 0) {
pblocknr = blocknr;
if (inode->i_ino != NILFS_DAT_INO) {
struct the_nilfs *nilfs = inode->i_sb->s_fs_info;
err = nilfs_dat_translate(nilfs->ns_dat, blocknr,
&pblocknr);
if (unlikely(err)) {
brelse(bh);
goto out_locked;
}
}
}
if (mode == READA) {
if (pblocknr != *submit_ptr + 1 || !trylock_buffer(bh)) {
err = -EBUSY;
brelse(bh);
goto out_locked;
}
} else {
lock_buffer(bh);
}
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
err = -EEXIST;
goto found;
}
set_buffer_mapped(bh);
bh->b_bdev = inode->i_sb->s_bdev;
bh->b_blocknr = pblocknr;
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(mode, bh);
bh->b_blocknr = blocknr;
*submit_ptr = pblocknr;
err = 0;
found:
*pbh = bh;
out_locked:
unlock_page(page);
page_cache_release(page);
return err;
}
int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
sector_t pblocknr, struct buffer_head **pbh,
int newblk)
{
struct buffer_head *bh;
struct inode *inode = NILFS_BTNC_I(btnc);
int err;
btnode_debug(3, "called: blocknr=%llu pblocknr=%llu new=%d ino=%lu\n",
(unsigned long long)blocknr, (unsigned long long)pblocknr,
newblk, inode->i_ino);
bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node);
if (unlikely(!bh))
return -ENOMEM;
err = -EEXIST; /* internal code */
if (newblk) {
if (unlikely(buffer_mapped(bh) || buffer_uptodate(bh) ||
buffer_dirty(bh))) {
BH_DEBUG(bh, "invalid new bh");
BUG();
}
bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev;
bh->b_blocknr = blocknr;
set_buffer_mapped(bh);
set_buffer_uptodate(bh);
goto found;
}
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);
btnode_debug(1, "return %d (xlate).\n", err);
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);
btnode_debug(3, "done (err=%d)\n", err);
return err;
}
int journal_write_metadata_buffer(transaction_t *transaction,
struct journal_head *jh_in,
struct journal_head **jh_out,
unsigned int blocknr)
{
int need_copy_out = 0;
int done_copy_out = 0;
int do_escape = 0;
char *mapped_data;
struct buffer_head *new_bh;
struct journal_head *new_jh;
struct page *new_page;
unsigned int new_offset;
struct buffer_head *bh_in = jh2bh(jh_in);
journal_t *journal = transaction->t_journal;
/*
* The buffer really shouldn't be locked: only the current committing
* transaction is allowed to write it, so nobody else is allowed
* to do any IO.
*
* akpm: except if we're journalling data, and write() output is
* also part of a shared mapping, and another thread has
* decided to launch a writepage() against this buffer.
*/
J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in));
new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL);
/* keep subsequent assertions sane */
new_bh->b_state = 0;
init_buffer(new_bh, NULL, NULL);
atomic_set(&new_bh->b_count, 1);
new_jh = journal_add_journal_head(new_bh); /* This sleeps */
/*
* If a new transaction has already done a buffer copy-out, then
* we use that version of the data for the commit.
*/
jbd_lock_bh_state(bh_in);
repeat:
if (jh_in->b_frozen_data) {
done_copy_out = 1;
new_page = virt_to_page(jh_in->b_frozen_data);
new_offset = offset_in_page(jh_in->b_frozen_data);
} else {
new_page = jh2bh(jh_in)->b_page;
new_offset = offset_in_page(jh2bh(jh_in)->b_data);
}
mapped_data = kmap_atomic(new_page);
/*
* Check for escaping
*/
if (*((__be32 *)(mapped_data + new_offset)) ==
cpu_to_be32(JFS_MAGIC_NUMBER)) {
need_copy_out = 1;
do_escape = 1;
}
kunmap_atomic(mapped_data);
/*
* Do we need to do a data copy?
*/
if (need_copy_out && !done_copy_out) {
char *tmp;
jbd_unlock_bh_state(bh_in);
tmp = jbd_alloc(bh_in->b_size, GFP_NOFS);
jbd_lock_bh_state(bh_in);
if (jh_in->b_frozen_data) {
jbd_free(tmp, bh_in->b_size);
goto repeat;
}
jh_in->b_frozen_data = tmp;
mapped_data = kmap_atomic(new_page);
memcpy(tmp, mapped_data + new_offset, jh2bh(jh_in)->b_size);
kunmap_atomic(mapped_data);
new_page = virt_to_page(tmp);
new_offset = offset_in_page(tmp);
done_copy_out = 1;
}
/*
* Did we need to do an escaping? Now we've done all the
* copying, we can finally do so.
*/
if (do_escape) {
mapped_data = kmap_atomic(new_page);
*((unsigned int *)(mapped_data + new_offset)) = 0;
kunmap_atomic(mapped_data);
}
set_bh_page(new_bh, new_page, new_offset);
new_jh->b_transaction = NULL;
new_bh->b_size = jh2bh(jh_in)->b_size;
new_bh->b_bdev = transaction->t_journal->j_dev;
new_bh->b_blocknr = blocknr;
set_buffer_mapped(new_bh);
set_buffer_dirty(new_bh);
*jh_out = new_jh;
/*
* The to-be-written buffer needs to get moved to the io queue,
* and the original buffer whose contents we are shadowing or
* copying is moved to the transaction's shadow queue.
*/
JBUFFER_TRACE(jh_in, "file as BJ_Shadow");
spin_lock(&journal->j_list_lock);
__journal_file_buffer(jh_in, transaction, BJ_Shadow);
spin_unlock(&journal->j_list_lock);
jbd_unlock_bh_state(bh_in);
JBUFFER_TRACE(new_jh, "file as BJ_IO");
journal_file_buffer(new_jh, transaction, BJ_IO);
return do_escape | (done_copy_out << 1);
}
int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr,
sector_t pblocknr, int mode, int mode_flags,
struct buffer_head **pbh, sector_t *submit_ptr)
{
struct buffer_head *bh;
struct inode *inode = NILFS_BTNC_I(btnc);
struct page *page;
int err;
bh = nilfs_grab_buffer(inode, btnc, blocknr, BIT(BH_NILFS_Node));
if (unlikely(!bh))
return -ENOMEM;
err = -EEXIST; /* internal code */
page = bh->b_page;
if (buffer_uptodate(bh) || buffer_dirty(bh))
goto found;
if (pblocknr == 0) {
pblocknr = blocknr;
if (inode->i_ino != NILFS_DAT_INO) {
struct the_nilfs *nilfs = inode->i_sb->s_fs_info;
/* blocknr is a virtual block number */
err = nilfs_dat_translate(nilfs->ns_dat, blocknr,
&pblocknr);
if (unlikely(err)) {
brelse(bh);
goto out_locked;
}
}
}
if (mode_flags & REQ_RAHEAD) {
if (pblocknr != *submit_ptr + 1 || !trylock_buffer(bh)) {
err = -EBUSY; /* internal code */
brelse(bh);
goto out_locked;
}
} else { /* mode == READ */
lock_buffer(bh);
}
if (buffer_uptodate(bh)) {
unlock_buffer(bh);
err = -EEXIST; /* internal code */
goto found;
}
set_buffer_mapped(bh);
bh->b_bdev = inode->i_sb->s_bdev;
bh->b_blocknr = pblocknr; /* set block address for read */
bh->b_end_io = end_buffer_read_sync;
get_bh(bh);
submit_bh(mode, mode_flags, bh);
bh->b_blocknr = blocknr; /* set back to the given block address */
*submit_ptr = pblocknr;
err = 0;
found:
*pbh = bh;
out_locked:
unlock_page(page);
put_page(page);
return err;
}
/* read a page from a file.
* We both read the page, and attach buffers to the page to record the
* address of each block (using bmap). These addresses will be used
* to write the block later, completely bypassing the filesystem.
* This usage is similar to how swap files are handled, and allows us
* to write to a file with no concerns of memory allocation failing.
*/
static struct page *read_page(struct file *file, unsigned long index,
struct bitmap *bitmap,
unsigned long count)
{
struct page *page = NULL;
struct inode *inode = file->f_path.dentry->d_inode;
struct buffer_head *bh;
sector_t block;
PRINTK("read bitmap file (%dB @ %Lu)\n", (int)PAGE_SIZE,
(unsigned long long)index << PAGE_SHIFT);
page = alloc_page(GFP_KERNEL);
if (!page)
page = ERR_PTR(-ENOMEM);
if (IS_ERR(page))
goto out;
bh = alloc_page_buffers(page, 1<<inode->i_blkbits, 0);
if (!bh) {
put_page(page);
page = ERR_PTR(-ENOMEM);
goto out;
}
attach_page_buffers(page, bh);
block = index << (PAGE_SHIFT - inode->i_blkbits);
while (bh) {
if (count == 0)
bh->b_blocknr = 0;
else {
bh->b_blocknr = bmap(inode, block);
if (bh->b_blocknr == 0) {
/* Cannot use this file! */
free_buffers(page);
page = ERR_PTR(-EINVAL);
goto out;
}
bh->b_bdev = inode->i_sb->s_bdev;
if (count < (1<<inode->i_blkbits))
count = 0;
else
count -= (1<<inode->i_blkbits);
bh->b_end_io = end_bitmap_write;
bh->b_private = bitmap;
atomic_inc(&bitmap->pending_writes);
set_buffer_locked(bh);
set_buffer_mapped(bh);
submit_bh(READ, bh);
}
block++;
bh = bh->b_this_page;
}
page->index = index;
wait_event(bitmap->write_wait,
atomic_read(&bitmap->pending_writes)==0);
if (bitmap->flags & BITMAP_WRITE_ERROR) {
free_buffers(page);
page = ERR_PTR(-EIO);
}
out:
if (IS_ERR(page))
printk(KERN_ALERT "md: bitmap read error: (%dB @ %Lu): %ld\n",
(int)PAGE_SIZE,
(unsigned long long)index << PAGE_SHIFT,
PTR_ERR(page));
return page;
}
/*
* 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;
#if defined(YANQIN)
struct radix_tree_root *rtree = &(NILFS_I(inode)->i_gc_blocks);
struct nilfs_gc_block_info *gbi;
#endif
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;
#if defined(YANQIN)
set_bit(BH_PrivateStart, &bh->b_state);
gbi = kmalloc(sizeof(struct nilfs_gc_block_info), GFP_NOFS);
if (!gbi) return -ENOMEM;
gbi->old_pblocknr = pbn;
gbi->new_pblocknr = 0; /* To be assigned later */
err = radix_tree_insert(rtree, vbn, gbi);
#endif
failed:
unlock_page(bh->b_page);
page_cache_release(bh->b_page);
return err;
}
