Eric Lee / linux-smarc-t335x-v3.2

Commit 87e99511ea54510ffb60b98001d108794d5037f8

Authored by Christoph Hellwig 2010-08-11 23:05:45 +0800

Committed by Al Viro 2010-08-18 13:09:00 +0800

Exists in master and in 4 other branches

kill BH_Ordered flag

Instead of abusing a buffer_head flag just add a variant of
sync_dirty_buffer which allows passing the exact type of write
flag required.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

Showing 5 changed files with 63 additions and 73 deletions Inline Diff

fs/buffer.c
fs/jbd/commit.c
fs/jbd2/commit.c
fs/nilfs2/super.c
include/linux/buffer_head.h

fs/buffer.c

Diff comments View file @ 87e9951

 /*
  *  linux/fs/buffer.c
  *
  *  Copyright (C) 1991, 1992, 2002  Linus Torvalds
  */
 /*
  * Start bdflush() with kernel_thread not syscall - Paul Gortmaker, 12/95
  *
  * Removed a lot of unnecessary code and simplified things now that
  * the buffer cache isn't our primary cache - Andrew Tridgell 12/96
  *
  * Speed up hash, lru, and free list operations.  Use gfp() for allocating
  * hash table, use SLAB cache for buffer heads. SMP threading.  -DaveM
  *
  * Added 32k buffer block sizes - these are required older ARM systems. - RMK
  *
  * async buffer flushing, 1999 Andrea Arcangeli <andrea@suse.de>
  */
 #include <linux/kernel.h>
 #include <linux/syscalls.h>
 #include <linux/fs.h>
 #include <linux/mm.h>
 #include <linux/percpu.h>
 #include <linux/slab.h>
 #include <linux/capability.h>
 #include <linux/blkdev.h>
 #include <linux/file.h>
 #include <linux/quotaops.h>
 #include <linux/highmem.h>
 #include <linux/module.h>
 #include <linux/writeback.h>
 #include <linux/hash.h>
 #include <linux/suspend.h>
 #include <linux/buffer_head.h>
 #include <linux/task_io_accounting_ops.h>
 #include <linux/bio.h>
 #include <linux/notifier.h>
 #include <linux/cpu.h>
 #include <linux/bitops.h>
 #include <linux/mpage.h>
 #include <linux/bit_spinlock.h>
 static int fsync_buffers_list(spinlock_t *lock, struct list_head *list);
 #define BH_ENTRY(list) list_entry((list), struct buffer_head, b_assoc_buffers)
 inline void
 init_buffer(struct buffer_head *bh, bh_end_io_t *handler, void *private)
 {
 	bh->b_end_io = handler;
 	bh->b_private = private;
 }
 EXPORT_SYMBOL(init_buffer);
 static int sync_buffer(void *word)
 {
 	struct block_device *bd;
 	struct buffer_head *bh
 		= container_of(word, struct buffer_head, b_state);
 	smp_mb();
 	bd = bh->b_bdev;
 	if (bd)
 		blk_run_address_space(bd->bd_inode->i_mapping);
 	io_schedule();
 	return 0;
 }
 void __lock_buffer(struct buffer_head *bh)
 {
 	wait_on_bit_lock(&bh->b_state, BH_Lock, sync_buffer,
 							TASK_UNINTERRUPTIBLE);
 }
 EXPORT_SYMBOL(__lock_buffer);
 void unlock_buffer(struct buffer_head *bh)
 {
 	clear_bit_unlock(BH_Lock, &bh->b_state);
 	smp_mb__after_clear_bit();
 	wake_up_bit(&bh->b_state, BH_Lock);
 }
 EXPORT_SYMBOL(unlock_buffer);
 /*
  * Block until a buffer comes unlocked.  This doesn't stop it
  * from becoming locked again - you have to lock it yourself
  * if you want to preserve its state.
  */
 void __wait_on_buffer(struct buffer_head * bh)
 {
 	wait_on_bit(&bh->b_state, BH_Lock, sync_buffer, TASK_UNINTERRUPTIBLE);
 }
 EXPORT_SYMBOL(__wait_on_buffer);
 static void
 __clear_page_buffers(struct page *page)
 {
 	ClearPagePrivate(page);
 	set_page_private(page, 0);
 	page_cache_release(page);
 }
 static int quiet_error(struct buffer_head *bh)
 {
 	if (!test_bit(BH_Quiet, &bh->b_state) && printk_ratelimit())
 		return 0;
 	return 1;
 }
 static void buffer_io_error(struct buffer_head *bh)
 {
 	char b[BDEVNAME_SIZE];
 	printk(KERN_ERR "Buffer I/O error on device %s, logical block %Lu\n",
 			bdevname(bh->b_bdev, b),
 			(unsigned long long)bh->b_blocknr);
 }
 /*
  * End-of-IO handler helper function which does not touch the bh after
  * unlocking it.
  * Note: unlock_buffer() sort-of does touch the bh after unlocking it, but
  * a race there is benign: unlock_buffer() only use the bh's address for
  * hashing after unlocking the buffer, so it doesn't actually touch the bh
  * itself.
  */
 static void __end_buffer_read_notouch(struct buffer_head *bh, int uptodate)
 {
 	if (uptodate) {
 		set_buffer_uptodate(bh);
 	} else {
 		/* This happens, due to failed READA attempts. */
 		clear_buffer_uptodate(bh);
 	}
 	unlock_buffer(bh);
 }
 /*
  * Default synchronous end-of-IO handler..  Just mark it up-to-date and
  * unlock the buffer. This is what ll_rw_block uses too.
  */
 void end_buffer_read_sync(struct buffer_head *bh, int uptodate)
 {
 	__end_buffer_read_notouch(bh, uptodate);
 	put_bh(bh);
 }
 EXPORT_SYMBOL(end_buffer_read_sync);
 void end_buffer_write_sync(struct buffer_head *bh, int uptodate)
 {
 	char b[BDEVNAME_SIZE];
 	if (uptodate) {
 		set_buffer_uptodate(bh);
 	} else {
 		if (!buffer_eopnotsupp(bh) && !quiet_error(bh)) {
 			buffer_io_error(bh);
 			printk(KERN_WARNING "lost page write due to "
 					"I/O error on %s\n",
 				       bdevname(bh->b_bdev, b));
 		}
 		set_buffer_write_io_error(bh);
 		clear_buffer_uptodate(bh);
 	}
 	unlock_buffer(bh);
 	put_bh(bh);
 }
 EXPORT_SYMBOL(end_buffer_write_sync);
 /*
  * Various filesystems appear to want __find_get_block to be non-blocking.
  * But it's the page lock which protects the buffers.  To get around this,
  * we get exclusion from try_to_free_buffers with the blockdev mapping's
  * private_lock.
  *
  * Hack idea: for the blockdev mapping, i_bufferlist_lock contention
  * may be quite high.  This code could TryLock the page, and if that
  * succeeds, there is no need to take private_lock. (But if
  * private_lock is contended then so is mapping->tree_lock).
  */
 static struct buffer_head *
 __find_get_block_slow(struct block_device *bdev, sector_t block)
 {
 	struct inode *bd_inode = bdev->bd_inode;
 	struct address_space *bd_mapping = bd_inode->i_mapping;
 	struct buffer_head *ret = NULL;
 	pgoff_t index;
 	struct buffer_head *bh;
 	struct buffer_head *head;
 	struct page *page;
 	int all_mapped = 1;
 	index = block >> (PAGE_CACHE_SHIFT - bd_inode->i_blkbits);
 	page = find_get_page(bd_mapping, index);
 	if (!page)
 		goto out;
 	spin_lock(&bd_mapping->private_lock);
 	if (!page_has_buffers(page))
 		goto out_unlock;
 	head = page_buffers(page);
 	bh = head;
 	do {
 		if (!buffer_mapped(bh))
 			all_mapped = 0;
 		else if (bh->b_blocknr == block) {
 			ret = bh;
 			get_bh(bh);
 			goto out_unlock;
 		}
 		bh = bh->b_this_page;
 	} while (bh != head);
 	/* we might be here because some of the buffers on this page are
 	 * not mapped.  This is due to various races between
 	 * file io on the block device and getblk.  It gets dealt with
 	 * elsewhere, don't buffer_error if we had some unmapped buffers
 	 */
 	if (all_mapped) {
 		printk("__find_get_block_slow() failed. "
 			"block=%llu, b_blocknr=%llu\n",
 			(unsigned long long)block,
 			(unsigned long long)bh->b_blocknr);
 		printk("b_state=0x%08lx, b_size=%zu\n",
 			bh->b_state, bh->b_size);
 		printk("device blocksize: %d\n", 1 << bd_inode->i_blkbits);
 	}
 out_unlock:
 	spin_unlock(&bd_mapping->private_lock);
 	page_cache_release(page);
 out:
 	return ret;
 }
 /* If invalidate_buffers() will trash dirty buffers, it means some kind
    of fs corruption is going on. Trashing dirty data always imply losing
    information that was supposed to be just stored on the physical layer
    by the user.
    Thus invalidate_buffers in general usage is not allwowed to trash
    dirty buffers. For example ioctl(FLSBLKBUF) expects dirty data to
    be preserved.  These buffers are simply skipped.
    We also skip buffers which are still in use.  For example this can
    happen if a userspace program is reading the block device.
    NOTE: In the case where the user removed a removable-media-disk even if
    there's still dirty data not synced on disk (due a bug in the device driver
    or due an error of the user), by not destroying the dirty buffers we could
    generate corruption also on the next media inserted, thus a parameter is
    necessary to handle this case in the most safe way possible (trying
    to not corrupt also the new disk inserted with the data belonging to
    the old now corrupted disk). Also for the ramdisk the natural thing
    to do in order to release the ramdisk memory is to destroy dirty buffers.
    These are two special cases. Normal usage imply the device driver
    to issue a sync on the device (without waiting I/O completion) and
    then an invalidate_buffers call that doesn't trash dirty buffers.
    For handling cache coherency with the blkdev pagecache the 'update' case
    is been introduced. It is needed to re-read from disk any pinned
    buffer. NOTE: re-reading from disk is destructive so we can do it only
    when we assume nobody is changing the buffercache under our I/O and when
    we think the disk contains more recent information than the buffercache.
    The update == 1 pass marks the buffers we need to update, the update == 2
    pass does the actual I/O. */
 void invalidate_bdev(struct block_device *bdev)
 {
 	struct address_space *mapping = bdev->bd_inode->i_mapping;
 	if (mapping->nrpages == 0)
 		return;
 	invalidate_bh_lrus();
 	lru_add_drain_all();	/* make sure all lru add caches are flushed */
 	invalidate_mapping_pages(mapping, 0, -1);
 }
 EXPORT_SYMBOL(invalidate_bdev);
 /*
  * Kick the writeback threads then try to free up some ZONE_NORMAL memory.
  */
 static void free_more_memory(void)
 {
 	struct zone *zone;
 	int nid;
 	wakeup_flusher_threads(1024);
 	yield();
 	for_each_online_node(nid) {
 		(void)first_zones_zonelist(node_zonelist(nid, GFP_NOFS),
 						gfp_zone(GFP_NOFS), NULL,
 						&zone);
 		if (zone)
 			try_to_free_pages(node_zonelist(nid, GFP_NOFS), 0,
 						GFP_NOFS, NULL);
 	}
 }
 /*
  * I/O completion handler for block_read_full_page() - pages
  * which come unlocked at the end of I/O.
  */
 static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
 {
 	unsigned long flags;
 	struct buffer_head *first;
 	struct buffer_head *tmp;
 	struct page *page;
 	int page_uptodate = 1;
 	BUG_ON(!buffer_async_read(bh));
 	page = bh->b_page;
 	if (uptodate) {
 		set_buffer_uptodate(bh);
 	} else {
 		clear_buffer_uptodate(bh);
 		if (!quiet_error(bh))
 			buffer_io_error(bh);
 		SetPageError(page);
 	}
 	/*
 	 * Be _very_ careful from here on. Bad things can happen if
 	 * two buffer heads end IO at almost the same time and both
 	 * decide that the page is now completely done.
 	 */
 	first = page_buffers(page);
 	local_irq_save(flags);
 	bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
 	clear_buffer_async_read(bh);
 	unlock_buffer(bh);
 	tmp = bh;
 	do {
 		if (!buffer_uptodate(tmp))
 			page_uptodate = 0;
 		if (buffer_async_read(tmp)) {
 			BUG_ON(!buffer_locked(tmp));
 			goto still_busy;
 		}
 		tmp = tmp->b_this_page;
 	} while (tmp != bh);
 	bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
 	local_irq_restore(flags);
 	/*
 	 * If none of the buffers had errors and they are all
 	 * uptodate then we can set the page uptodate.
 	 */
 	if (page_uptodate && !PageError(page))
 		SetPageUptodate(page);
 	unlock_page(page);
 	return;
 still_busy:
 	bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
 	local_irq_restore(flags);
 	return;
 }
 /*
  * Completion handler for block_write_full_page() - pages which are unlocked
  * during I/O, and which have PageWriteback cleared upon I/O completion.
  */
 void end_buffer_async_write(struct buffer_head *bh, int uptodate)
 {
 	char b[BDEVNAME_SIZE];
 	unsigned long flags;
 	struct buffer_head *first;
 	struct buffer_head *tmp;
 	struct page *page;
 	BUG_ON(!buffer_async_write(bh));
 	page = bh->b_page;
 	if (uptodate) {
 		set_buffer_uptodate(bh);
 	} else {
 		if (!quiet_error(bh)) {
 			buffer_io_error(bh);
 			printk(KERN_WARNING "lost page write due to "
 					"I/O error on %s\n",
 			       bdevname(bh->b_bdev, b));
 		}
 		set_bit(AS_EIO, &page->mapping->flags);
 		set_buffer_write_io_error(bh);
 		clear_buffer_uptodate(bh);
 		SetPageError(page);
 	}
 	first = page_buffers(page);
 	local_irq_save(flags);
 	bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
 	clear_buffer_async_write(bh);
 	unlock_buffer(bh);
 	tmp = bh->b_this_page;
 	while (tmp != bh) {
 		if (buffer_async_write(tmp)) {
 			BUG_ON(!buffer_locked(tmp));
 			goto still_busy;
 		}
 		tmp = tmp->b_this_page;
 	}
 	bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
 	local_irq_restore(flags);
 	end_page_writeback(page);
 	return;
 still_busy:
 	bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
 	local_irq_restore(flags);
 	return;
 }
 EXPORT_SYMBOL(end_buffer_async_write);
 /*
  * If a page's buffers are under async readin (end_buffer_async_read
  * completion) then there is a possibility that another thread of
  * control could lock one of the buffers after it has completed
  * but while some of the other buffers have not completed.  This
  * locked buffer would confuse end_buffer_async_read() into not unlocking
  * the page.  So the absence of BH_Async_Read tells end_buffer_async_read()
  * that this buffer is not under async I/O.
  *
  * The page comes unlocked when it has no locked buffer_async buffers
  * left.
  *
  * PageLocked prevents anyone starting new async I/O reads any of
  * the buffers.
  *
  * PageWriteback is used to prevent simultaneous writeout of the same
  * page.
  *
  * PageLocked prevents anyone from starting writeback of a page which is
  * under read I/O (PageWriteback is only ever set against a locked page).
  */
 static void mark_buffer_async_read(struct buffer_head *bh)
 {
 	bh->b_end_io = end_buffer_async_read;
 	set_buffer_async_read(bh);
 }
 static void mark_buffer_async_write_endio(struct buffer_head *bh,
 					  bh_end_io_t *handler)
 {
 	bh->b_end_io = handler;
 	set_buffer_async_write(bh);
 }
 void mark_buffer_async_write(struct buffer_head *bh)
 {
 	mark_buffer_async_write_endio(bh, end_buffer_async_write);
 }
 EXPORT_SYMBOL(mark_buffer_async_write);
 /*
  * fs/buffer.c contains helper functions for buffer-backed address space's
  * fsync functions.  A common requirement for buffer-based filesystems is
  * that certain data from the backing blockdev needs to be written out for
  * a successful fsync().  For example, ext2 indirect blocks need to be
  * written back and waited upon before fsync() returns.
  *
  * The functions mark_buffer_inode_dirty(), fsync_inode_buffers(),
  * inode_has_buffers() and invalidate_inode_buffers() are provided for the
  * management of a list of dependent buffers at ->i_mapping->private_list.
  *
  * Locking is a little subtle: try_to_free_buffers() will remove buffers
  * from their controlling inode's queue when they are being freed.  But
  * try_to_free_buffers() will be operating against the *blockdev* mapping
  * at the time, not against the S_ISREG file which depends on those buffers.
  * So the locking for private_list is via the private_lock in the address_space
  * which backs the buffers.  Which is different from the address_space
  * against which the buffers are listed.  So for a particular address_space,
  * mapping->private_lock does *not* protect mapping->private_list!  In fact,
  * mapping->private_list will always be protected by the backing blockdev's
  * ->private_lock.
  *
  * Which introduces a requirement: all buffers on an address_space's
  * ->private_list must be from the same address_space: the blockdev's.
  *
  * address_spaces which do not place buffers at ->private_list via these
  * utility functions are free to use private_lock and private_list for
  * whatever they want.  The only requirement is that list_empty(private_list)
  * be true at clear_inode() time.
  *
  * FIXME: clear_inode should not call invalidate_inode_buffers().  The
  * filesystems should do that.  invalidate_inode_buffers() should just go
  * BUG_ON(!list_empty).
  *
  * FIXME: mark_buffer_dirty_inode() is a data-plane operation.  It should
  * take an address_space, not an inode.  And it should be called
  * mark_buffer_dirty_fsync() to clearly define why those buffers are being
  * queued up.
  *
  * FIXME: mark_buffer_dirty_inode() doesn't need to add the buffer to the
  * list if it is already on a list.  Because if the buffer is on a list,
  * it *must* already be on the right one.  If not, the filesystem is being
  * silly.  This will save a ton of locking.  But first we have to ensure
  * that buffers are taken *off* the old inode's list when they are freed
  * (presumably in truncate).  That requires careful auditing of all
  * filesystems (do it inside bforget()).  It could also be done by bringing
  * b_inode back.
  */
 /*
  * The buffer's backing address_space's private_lock must be held
  */
 static void __remove_assoc_queue(struct buffer_head *bh)
 {
 	list_del_init(&bh->b_assoc_buffers);
 	WARN_ON(!bh->b_assoc_map);
 	if (buffer_write_io_error(bh))
 		set_bit(AS_EIO, &bh->b_assoc_map->flags);
 	bh->b_assoc_map = NULL;
 }
 int inode_has_buffers(struct inode *inode)
 {
 	return !list_empty(&inode->i_data.private_list);
 }
 /*
  * osync is designed to support O_SYNC io.  It waits synchronously for
  * all already-submitted IO to complete, but does not queue any new
  * writes to the disk.
  *
  * To do O_SYNC writes, just queue the buffer writes with ll_rw_block as
  * you dirty the buffers, and then use osync_inode_buffers to wait for
  * completion.  Any other dirty buffers which are not yet queued for
  * write will not be flushed to disk by the osync.
  */
 static int osync_buffers_list(spinlock_t *lock, struct list_head *list)
 {
 	struct buffer_head *bh;
 	struct list_head *p;
 	int err = 0;
 	spin_lock(lock);
 repeat:
 	list_for_each_prev(p, list) {
 		bh = BH_ENTRY(p);
 		if (buffer_locked(bh)) {
 			get_bh(bh);
 			spin_unlock(lock);
 			wait_on_buffer(bh);
 			if (!buffer_uptodate(bh))
 				err = -EIO;
 			brelse(bh);
 			spin_lock(lock);
 			goto repeat;
 		}
 	}
 	spin_unlock(lock);
 	return err;
 }
 static void do_thaw_one(struct super_block *sb, void *unused)
 {
 	char b[BDEVNAME_SIZE];
 	while (sb->s_bdev && !thaw_bdev(sb->s_bdev, sb))
 		printk(KERN_WARNING "Emergency Thaw on %s\n",
 		       bdevname(sb->s_bdev, b));
 }
 static void do_thaw_all(struct work_struct *work)
 {
 	iterate_supers(do_thaw_one, NULL);
 	kfree(work);
 	printk(KERN_WARNING "Emergency Thaw complete\n");
 }
 /**
  * emergency_thaw_all -- forcibly thaw every frozen filesystem
  *
  * Used for emergency unfreeze of all filesystems via SysRq
  */
 void emergency_thaw_all(void)
 {
 	struct work_struct *work;
 	work = kmalloc(sizeof(*work), GFP_ATOMIC);
 	if (work) {
 		INIT_WORK(work, do_thaw_all);
 		schedule_work(work);
 	}
 }
 /**
  * sync_mapping_buffers - write out & wait upon a mapping's "associated" buffers
  * @mapping: the mapping which wants those buffers written
  *
  * Starts I/O against the buffers at mapping->private_list, and waits upon
  * that I/O.
  *
  * Basically, this is a convenience function for fsync().
  * @mapping is a file or directory which needs those buffers to be written for
  * a successful fsync().
  */
 int sync_mapping_buffers(struct address_space *mapping)
 {
 	struct address_space *buffer_mapping = mapping->assoc_mapping;
 	if (buffer_mapping == NULL || list_empty(&mapping->private_list))
 		return 0;
 	return fsync_buffers_list(&buffer_mapping->private_lock,
 					&mapping->private_list);
 }
 EXPORT_SYMBOL(sync_mapping_buffers);
 /*
  * Called when we've recently written block `bblock', and it is known that
  * `bblock' was for a buffer_boundary() buffer.  This means that the block at
  * `bblock + 1' is probably a dirty indirect block.  Hunt it down and, if it's
  * dirty, schedule it for IO.  So that indirects merge nicely with their data.
  */
 void write_boundary_block(struct block_device *bdev,
 			sector_t bblock, unsigned blocksize)
 {
 	struct buffer_head *bh = __find_get_block(bdev, bblock + 1, blocksize);
 	if (bh) {
 		if (buffer_dirty(bh))
 			ll_rw_block(WRITE, 1, &bh);
 		put_bh(bh);
 	}
 }
 void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode)
 {
 	struct address_space *mapping = inode->i_mapping;
 	struct address_space *buffer_mapping = bh->b_page->mapping;
 	mark_buffer_dirty(bh);
 	if (!mapping->assoc_mapping) {
 		mapping->assoc_mapping = buffer_mapping;
 	} else {
 		BUG_ON(mapping->assoc_mapping != buffer_mapping);
 	}
 	if (!bh->b_assoc_map) {
 		spin_lock(&buffer_mapping->private_lock);
 		list_move_tail(&bh->b_assoc_buffers,
 				&mapping->private_list);
 		bh->b_assoc_map = mapping;
 		spin_unlock(&buffer_mapping->private_lock);
 	}
 }
 EXPORT_SYMBOL(mark_buffer_dirty_inode);
 /*
  * Mark the page dirty, and set it dirty in the radix tree, and mark the inode
  * dirty.
  *
  * If warn is true, then emit a warning if the page is not uptodate and has
  * not been truncated.
  */
 static void __set_page_dirty(struct page *page,
 		struct address_space *mapping, int warn)
 {
 	spin_lock_irq(&mapping->tree_lock);
 	if (page->mapping) {	/* Race with truncate? */
 		WARN_ON_ONCE(warn && !PageUptodate(page));
 		account_page_dirtied(page, mapping);
 		radix_tree_tag_set(&mapping->page_tree,
 				page_index(page), PAGECACHE_TAG_DIRTY);
 	}
 	spin_unlock_irq(&mapping->tree_lock);
 	__mark_inode_dirty(mapping->host, I_DIRTY_PAGES);
 }
 /*
  * Add a page to the dirty page list.
  *
  * It is a sad fact of life that this function is called from several places
  * deeply under spinlocking.  It may not sleep.
  *
  * If the page has buffers, the uptodate buffers are set dirty, to preserve
  * dirty-state coherency between the page and the buffers.  It the page does
  * not have buffers then when they are later attached they will all be set
  * dirty.
  *
  * The buffers are dirtied before the page is dirtied.  There's a small race
  * window in which a writepage caller may see the page cleanness but not the
  * buffer dirtiness.  That's fine.  If this code were to set the page dirty
  * before the buffers, a concurrent writepage caller could clear the page dirty
  * bit, see a bunch of clean buffers and we'd end up with dirty buffers/clean
  * page on the dirty page list.
  *
  * We use private_lock to lock against try_to_free_buffers while using the
  * page's buffer list.  Also use this to protect against clean buffers being
  * added to the page after it was set dirty.
  *
  * FIXME: may need to call ->reservepage here as well.  That's rather up to the
  * address_space though.
  */
 int __set_page_dirty_buffers(struct page *page)
 {
 	int newly_dirty;
 	struct address_space *mapping = page_mapping(page);
 	if (unlikely(!mapping))
 		return !TestSetPageDirty(page);
 	spin_lock(&mapping->private_lock);
 	if (page_has_buffers(page)) {
 		struct buffer_head *head = page_buffers(page);
 		struct buffer_head *bh = head;
 		do {
 			set_buffer_dirty(bh);
 			bh = bh->b_this_page;
 		} while (bh != head);
 	}
 	newly_dirty = !TestSetPageDirty(page);
 	spin_unlock(&mapping->private_lock);
 	if (newly_dirty)
 		__set_page_dirty(page, mapping, 1);
 	return newly_dirty;
 }
 EXPORT_SYMBOL(__set_page_dirty_buffers);
 /*
  * Write out and wait upon a list of buffers.
  *
  * We have conflicting pressures: we want to make sure that all
  * initially dirty buffers get waited on, but that any subsequently
  * dirtied buffers don't.  After all, we don't want fsync to last
  * forever if somebody is actively writing to the file.
  *
  * Do this in two main stages: first we copy dirty buffers to a
  * temporary inode list, queueing the writes as we go.  Then we clean
  * up, waiting for those writes to complete.
  *
  * During this second stage, any subsequent updates to the file may end
  * up refiling the buffer on the original inode's dirty list again, so
  * there is a chance we will end up with a buffer queued for write but
  * not yet completed on that list.  So, as a final cleanup we go through
  * the osync code to catch these locked, dirty buffers without requeuing
  * any newly dirty buffers for write.
  */
 static int fsync_buffers_list(spinlock_t *lock, struct list_head *list)
 {
 	struct buffer_head *bh;
 	struct list_head tmp;
 	struct address_space *mapping, *prev_mapping = NULL;
 	int err = 0, err2;
 	INIT_LIST_HEAD(&tmp);
 	spin_lock(lock);
 	while (!list_empty(list)) {
 		bh = BH_ENTRY(list->next);
 		mapping = bh->b_assoc_map;
 		__remove_assoc_queue(bh);
 		/* Avoid race with mark_buffer_dirty_inode() which does
 		 * a lockless check and we rely on seeing the dirty bit */
 		smp_mb();
 		if (buffer_dirty(bh) || buffer_locked(bh)) {
 			list_add(&bh->b_assoc_buffers, &tmp);
 			bh->b_assoc_map = mapping;
 			if (buffer_dirty(bh)) {
 				get_bh(bh);
 				spin_unlock(lock);
 				/*
 				 * Ensure any pending I/O completes so that
 				 * ll_rw_block() actually writes the current
 				 * contents - it is a noop if I/O is still in
 				 * flight on potentially older contents.
 				 */
 				ll_rw_block(SWRITE_SYNC_PLUG, 1, &bh);
 				/*
 				 * Kick off IO for the previous mapping. Note
 				 * that we will not run the very last mapping,
 				 * wait_on_buffer() will do that for us
 				 * through sync_buffer().
 				 */
 				if (prev_mapping && prev_mapping != mapping)
 					blk_run_address_space(prev_mapping);
 				prev_mapping = mapping;
 				brelse(bh);
 				spin_lock(lock);
 			}
 		}
 	}
 	while (!list_empty(&tmp)) {
 		bh = BH_ENTRY(tmp.prev);
 		get_bh(bh);
 		mapping = bh->b_assoc_map;
 		__remove_assoc_queue(bh);
 		/* Avoid race with mark_buffer_dirty_inode() which does
 		 * a lockless check and we rely on seeing the dirty bit */
 		smp_mb();
 		if (buffer_dirty(bh)) {
 			list_add(&bh->b_assoc_buffers,
 				 &mapping->private_list);
 			bh->b_assoc_map = mapping;
 		}
 		spin_unlock(lock);
 		wait_on_buffer(bh);
 		if (!buffer_uptodate(bh))
 			err = -EIO;
 		brelse(bh);
 		spin_lock(lock);
 	}
 	spin_unlock(lock);
 	err2 = osync_buffers_list(lock, list);
 	if (err)
 		return err;
 	else
 		return err2;
 }
 /*
  * Invalidate any and all dirty buffers on a given inode.  We are
  * probably unmounting the fs, but that doesn't mean we have already
  * done a sync().  Just drop the buffers from the inode list.
  *
  * NOTE: we take the inode's blockdev's mapping's private_lock.  Which
  * assumes that all the buffers are against the blockdev.  Not true
  * for reiserfs.
  */
 void invalidate_inode_buffers(struct inode *inode)
 {
 	if (inode_has_buffers(inode)) {
 		struct address_space *mapping = &inode->i_data;
 		struct list_head *list = &mapping->private_list;
 		struct address_space *buffer_mapping = mapping->assoc_mapping;
 		spin_lock(&buffer_mapping->private_lock);
 		while (!list_empty(list))
 			__remove_assoc_queue(BH_ENTRY(list->next));
 		spin_unlock(&buffer_mapping->private_lock);
 	}
 }
 EXPORT_SYMBOL(invalidate_inode_buffers);
 /*
  * Remove any clean buffers from the inode's buffer list.  This is called
  * when we're trying to free the inode itself.  Those buffers can pin it.
  *
  * Returns true if all buffers were removed.
  */
 int remove_inode_buffers(struct inode *inode)
 {
 	int ret = 1;
 	if (inode_has_buffers(inode)) {
 		struct address_space *mapping = &inode->i_data;
 		struct list_head *list = &mapping->private_list;
 		struct address_space *buffer_mapping = mapping->assoc_mapping;
 		spin_lock(&buffer_mapping->private_lock);
 		while (!list_empty(list)) {
 			struct buffer_head *bh = BH_ENTRY(list->next);
 			if (buffer_dirty(bh)) {
 				ret = 0;
 				break;
 			}
 			__remove_assoc_queue(bh);
 		}
 		spin_unlock(&buffer_mapping->private_lock);
 	}
 	return ret;
 }
 /*
  * Create the appropriate buffers when given a page for data area and
  * the size of each buffer.. Use the bh->b_this_page linked list to
  * follow the buffers created.  Return NULL if unable to create more
  * buffers.
  *
  * The retry flag is used to differentiate async IO (paging, swapping)
  * which may not fail from ordinary buffer allocations.
  */
 struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size,
 		int retry)
 {
 	struct buffer_head *bh, *head;
 	long offset;
 try_again:
 	head = NULL;
 	offset = PAGE_SIZE;
 	while ((offset -= size) >= 0) {
 		bh = alloc_buffer_head(GFP_NOFS);
 		if (!bh)
 			goto no_grow;
 		bh->b_bdev = NULL;
 		bh->b_this_page = head;
 		bh->b_blocknr = -1;
 		head = bh;
 		bh->b_state = 0;
 		atomic_set(&bh->b_count, 0);
 		bh->b_private = NULL;
 		bh->b_size = size;
 		/* Link the buffer to its page */
 		set_bh_page(bh, page, offset);
 		init_buffer(bh, NULL, NULL);
 	}
 	return head;
 /*
  * In case anything failed, we just free everything we got.
  */
 no_grow:
 	if (head) {
 		do {
 			bh = head;
 			head = head->b_this_page;
 			free_buffer_head(bh);
 		} while (head);
 	}
 	/*
 	 * Return failure for non-async IO requests.  Async IO requests
 	 * are not allowed to fail, so we have to wait until buffer heads
 	 * become available.  But we don't want tasks sleeping with
 	 * partially complete buffers, so all were released above.
 	 */
 	if (!retry)
 		return NULL;
 	/* We're _really_ low on memory. Now we just
 	 * wait for old buffer heads to become free due to
 	 * finishing IO.  Since this is an async request and
 	 * the reserve list is empty, we're sure there are
 	 * async buffer heads in use.
 	 */
 	free_more_memory();
 	goto try_again;
 }
 EXPORT_SYMBOL_GPL(alloc_page_buffers);
 static inline void
 link_dev_buffers(struct page *page, struct buffer_head *head)
 {
 	struct buffer_head *bh, *tail;
 	bh = head;
 	do {
 		tail = bh;
 		bh = bh->b_this_page;
 	} while (bh);
 	tail->b_this_page = head;
 	attach_page_buffers(page, head);
 }
 /*
  * Initialise the state of a blockdev page's buffers.
  */
 static void
 init_page_buffers(struct page *page, struct block_device *bdev,
 			sector_t block, int size)
 {
 	struct buffer_head *head = page_buffers(page);
 	struct buffer_head *bh = head;
 	int uptodate = PageUptodate(page);
 	do {
 		if (!buffer_mapped(bh)) {
 			init_buffer(bh, NULL, NULL);
 			bh->b_bdev = bdev;
 			bh->b_blocknr = block;
 			if (uptodate)
 				set_buffer_uptodate(bh);
 			set_buffer_mapped(bh);
 		}
 		block++;
 		bh = bh->b_this_page;
 	} while (bh != head);
 }
 /*
  * Create the page-cache page that contains the requested block.
  *
  * This is user purely for blockdev mappings.
  */
 static struct page *
 grow_dev_page(struct block_device *bdev, sector_t block,
 		pgoff_t index, int size)
 {
 	struct inode *inode = bdev->bd_inode;
 	struct page *page;
 	struct buffer_head *bh;
 	page = find_or_create_page(inode->i_mapping, index,
 		(mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS)|__GFP_MOVABLE);
 	if (!page)
 		return NULL;
 	BUG_ON(!PageLocked(page));
 	if (page_has_buffers(page)) {
 		bh = page_buffers(page);
 		if (bh->b_size == size) {
 			init_page_buffers(page, bdev, block, size);
 			return page;
 		}
 		if (!try_to_free_buffers(page))
 			goto failed;
 	}
 	/*
 	 * Allocate some buffers for this page
 	 */
 	bh = alloc_page_buffers(page, size, 0);
 	if (!bh)
 		goto failed;
 	/*
 	 * Link the page to the buffers and initialise them.  Take the
 	 * lock to be atomic wrt __find_get_block(), which does not
 	 * run under the page lock.
 	 */
 	spin_lock(&inode->i_mapping->private_lock);
 	link_dev_buffers(page, bh);
 	init_page_buffers(page, bdev, block, size);
 	spin_unlock(&inode->i_mapping->private_lock);
 	return page;
 failed:
 	BUG();
 	unlock_page(page);
 	page_cache_release(page);
 	return NULL;
 }
 /*
  * Create buffers for the specified block device block's page.  If
  * that page was dirty, the buffers are set dirty also.
  */
 static int
 grow_buffers(struct block_device *bdev, sector_t block, int size)
 {
 	struct page *page;
 	pgoff_t index;
 	int sizebits;
 	sizebits = -1;
 	do {
 		sizebits++;
 	} while ((size << sizebits) < PAGE_SIZE);
 	index = block >> sizebits;
 	/*
 	 * Check for a block which wants to lie outside our maximum possible
 	 * pagecache index.  (this comparison is done using sector_t types).
 	 */
 	if (unlikely(index != block >> sizebits)) {
 		char b[BDEVNAME_SIZE];
 		printk(KERN_ERR "%s: requested out-of-range block %llu for "
 			"device %s\n",
 			__func__, (unsigned long long)block,
 			bdevname(bdev, b));
 		return -EIO;
 	}
 	block = index << sizebits;
 	/* Create a page with the proper size buffers.. */
 	page = grow_dev_page(bdev, block, index, size);
 	if (!page)
 		return 0;
 	unlock_page(page);
 	page_cache_release(page);
 	return 1;
 }
 static struct buffer_head *
 __getblk_slow(struct block_device *bdev, sector_t block, int size)
 {
 	/* Size must be multiple of hard sectorsize */
 	if (unlikely(size & (bdev_logical_block_size(bdev)-1) ||
 			(size < 512 || size > PAGE_SIZE))) {
 		printk(KERN_ERR "getblk(): invalid block size %d requested\n",
 					size);
 		printk(KERN_ERR "logical block size: %d\n",
 					bdev_logical_block_size(bdev));
 		dump_stack();
 		return NULL;
 	}
 	for (;;) {
 		struct buffer_head * bh;
 		int ret;
 		bh = __find_get_block(bdev, block, size);
 		if (bh)
 			return bh;
 		ret = grow_buffers(bdev, block, size);
 		if (ret < 0)
 			return NULL;
 		if (ret == 0)
 			free_more_memory();
 	}
 }
 /*
  * The relationship between dirty buffers and dirty pages:
  *
  * Whenever a page has any dirty buffers, the page's dirty bit is set, and
  * the page is tagged dirty in its radix tree.
  *
  * At all times, the dirtiness of the buffers represents the dirtiness of
  * subsections of the page.  If the page has buffers, the page dirty bit is
  * merely a hint about the true dirty state.
  *
  * When a page is set dirty in its entirety, all its buffers are marked dirty
  * (if the page has buffers).
  *
  * When a buffer is marked dirty, its page is dirtied, but the page's other
  * buffers are not.
  *
  * Also.  When blockdev buffers are explicitly read with bread(), they
  * individually become uptodate.  But their backing page remains not
  * uptodate - even if all of its buffers are uptodate.  A subsequent
  * block_read_full_page() against that page will discover all the uptodate
  * buffers, will set the page uptodate and will perform no I/O.
  */
 /**
  * mark_buffer_dirty - mark a buffer_head as needing writeout
  * @bh: the buffer_head to mark dirty
  *
  * mark_buffer_dirty() will set the dirty bit against the buffer, then set its
  * backing page dirty, then tag the page as dirty in its address_space's radix
  * tree and then attach the address_space's inode to its superblock's dirty
  * inode list.
  *
  * mark_buffer_dirty() is atomic.  It takes bh->b_page->mapping->private_lock,
  * mapping->tree_lock and the global inode_lock.
  */
 void mark_buffer_dirty(struct buffer_head *bh)
 {
 	WARN_ON_ONCE(!buffer_uptodate(bh));
 	/*
 	 * Very *carefully* optimize the it-is-already-dirty case.
 	 *
 	 * Don't let the final "is it dirty" escape to before we
 	 * perhaps modified the buffer.
 	 */
 	if (buffer_dirty(bh)) {
 		smp_mb();
 		if (buffer_dirty(bh))
 			return;
 	}
 	if (!test_set_buffer_dirty(bh)) {
 		struct page *page = bh->b_page;
 		if (!TestSetPageDirty(page)) {
 			struct address_space *mapping = page_mapping(page);
 			if (mapping)
 				__set_page_dirty(page, mapping, 0);
 		}
 	}
 }
 EXPORT_SYMBOL(mark_buffer_dirty);
 /*
  * Decrement a buffer_head's reference count.  If all buffers against a page
  * have zero reference count, are clean and unlocked, and if the page is clean
  * and unlocked then try_to_free_buffers() may strip the buffers from the page
  * in preparation for freeing it (sometimes, rarely, buffers are removed from
  * a page but it ends up not being freed, and buffers may later be reattached).
  */
 void __brelse(struct buffer_head * buf)
 {
 	if (atomic_read(&buf->b_count)) {
 		put_bh(buf);
 		return;
 	}
 	WARN(1, KERN_ERR "VFS: brelse: Trying to free free buffer\n");
 }
 EXPORT_SYMBOL(__brelse);
 /*
  * bforget() is like brelse(), except it discards any
  * potentially dirty data.
  */
 void __bforget(struct buffer_head *bh)
 {
 	clear_buffer_dirty(bh);
 	if (bh->b_assoc_map) {
 		struct address_space *buffer_mapping = bh->b_page->mapping;
 		spin_lock(&buffer_mapping->private_lock);
 		list_del_init(&bh->b_assoc_buffers);
 		bh->b_assoc_map = NULL;
 		spin_unlock(&buffer_mapping->private_lock);
 	}
 	__brelse(bh);
 }
 EXPORT_SYMBOL(__bforget);
 static struct buffer_head *__bread_slow(struct buffer_head *bh)
 {
 	lock_buffer(bh);
 	if (buffer_uptodate(bh)) {
 		unlock_buffer(bh);
 		return bh;
 	} else {
 		get_bh(bh);
 		bh->b_end_io = end_buffer_read_sync;
 		submit_bh(READ, bh);
 		wait_on_buffer(bh);
 		if (buffer_uptodate(bh))
 			return bh;
 	}
 	brelse(bh);
 	return NULL;
 }
 /*
  * Per-cpu buffer LRU implementation.  To reduce the cost of __find_get_block().
  * The bhs[] array is sorted - newest buffer is at bhs[0].  Buffers have their
  * refcount elevated by one when they're in an LRU.  A buffer can only appear
  * once in a particular CPU's LRU.  A single buffer can be present in multiple
  * CPU's LRUs at the same time.
  *
  * This is a transparent caching front-end to sb_bread(), sb_getblk() and
  * sb_find_get_block().
  *
  * The LRUs themselves only need locking against invalidate_bh_lrus.  We use
  * a local interrupt disable for that.
  */
 #define BH_LRU_SIZE	8
 struct bh_lru {
 	struct buffer_head *bhs[BH_LRU_SIZE];
 };
 static DEFINE_PER_CPU(struct bh_lru, bh_lrus) = {{ NULL }};
 #ifdef CONFIG_SMP
 #define bh_lru_lock()	local_irq_disable()
 #define bh_lru_unlock()	local_irq_enable()
 #else
 #define bh_lru_lock()	preempt_disable()
 #define bh_lru_unlock()	preempt_enable()
 #endif
 static inline void check_irqs_on(void)
 {
 #ifdef irqs_disabled
 	BUG_ON(irqs_disabled());
 #endif
 }
 /*
  * The LRU management algorithm is dopey-but-simple.  Sorry.
  */
 static void bh_lru_install(struct buffer_head *bh)
 {
 	struct buffer_head *evictee = NULL;
 	struct bh_lru *lru;
 	check_irqs_on();
 	bh_lru_lock();
 	lru = &__get_cpu_var(bh_lrus);
 	if (lru->bhs[0] != bh) {
 		struct buffer_head *bhs[BH_LRU_SIZE];
 		int in;
 		int out = 0;
 		get_bh(bh);
 		bhs[out++] = bh;
 		for (in = 0; in < BH_LRU_SIZE; in++) {
 			struct buffer_head *bh2 = lru->bhs[in];
 			if (bh2 == bh) {
 				__brelse(bh2);
 			} else {
 				if (out >= BH_LRU_SIZE) {
 					BUG_ON(evictee != NULL);
 					evictee = bh2;
 				} else {
 					bhs[out++] = bh2;
 				}
 			}
 		}
 		while (out < BH_LRU_SIZE)
 			bhs[out++] = NULL;
 		memcpy(lru->bhs, bhs, sizeof(bhs));
 	}
 	bh_lru_unlock();
 	if (evictee)
 		__brelse(evictee);
 }
 /*
  * Look up the bh in this cpu's LRU.  If it's there, move it to the head.
  */
 static struct buffer_head *
 lookup_bh_lru(struct block_device *bdev, sector_t block, unsigned size)
 {
 	struct buffer_head *ret = NULL;
 	struct bh_lru *lru;
 	unsigned int i;
 	check_irqs_on();
 	bh_lru_lock();
 	lru = &__get_cpu_var(bh_lrus);
 	for (i = 0; i < BH_LRU_SIZE; i++) {
 		struct buffer_head *bh = lru->bhs[i];
 		if (bh && bh->b_bdev == bdev &&
 				bh->b_blocknr == block && bh->b_size == size) {
 			if (i) {
 				while (i) {
 					lru->bhs[i] = lru->bhs[i - 1];
 					i--;
 				}
 				lru->bhs[0] = bh;
 			}
 			get_bh(bh);
 			ret = bh;
 			break;
 		}
 	}
 	bh_lru_unlock();
 	return ret;
 }
 /*
  * Perform a pagecache lookup for the matching buffer.  If it's there, refresh
  * it in the LRU and mark it as accessed.  If it is not present then return
  * NULL
  */
 struct buffer_head *
 __find_get_block(struct block_device *bdev, sector_t block, unsigned size)
 {
 	struct buffer_head *bh = lookup_bh_lru(bdev, block, size);
 	if (bh == NULL) {
 		bh = __find_get_block_slow(bdev, block);
 		if (bh)
 			bh_lru_install(bh);
 	}
 	if (bh)
 		touch_buffer(bh);
 	return bh;
 }
 EXPORT_SYMBOL(__find_get_block);
 /*
  * __getblk will locate (and, if necessary, create) the buffer_head
  * which corresponds to the passed block_device, block and size. The
  * returned buffer has its reference count incremented.
  *
  * __getblk() cannot fail - it just keeps trying.  If you pass it an
  * illegal block number, __getblk() will happily return a buffer_head
  * which represents the non-existent block.  Very weird.
  *
  * __getblk() will lock up the machine if grow_dev_page's try_to_free_buffers()
  * attempt is failing.  FIXME, perhaps?
  */
 struct buffer_head *
 __getblk(struct block_device *bdev, sector_t block, unsigned size)
 {
 	struct buffer_head *bh = __find_get_block(bdev, block, size);
 	might_sleep();
 	if (bh == NULL)
 		bh = __getblk_slow(bdev, block, size);
 	return bh;
 }
 EXPORT_SYMBOL(__getblk);
 /*
  * Do async read-ahead on a buffer..
  */
 void __breadahead(struct block_device *bdev, sector_t block, unsigned size)
 {
 	struct buffer_head *bh = __getblk(bdev, block, size);
 	if (likely(bh)) {
 		ll_rw_block(READA, 1, &bh);
 		brelse(bh);
 	}
 }
 EXPORT_SYMBOL(__breadahead);
 /**
  *  __bread() - reads a specified block and returns the bh
  *  @bdev: the block_device to read from
  *  @block: number of block
  *  @size: size (in bytes) to read
  *
  *  Reads a specified block, and returns buffer head that contains it.
  *  It returns NULL if the block was unreadable.
  */
 struct buffer_head *
 __bread(struct block_device *bdev, sector_t block, unsigned size)
 {
 	struct buffer_head *bh = __getblk(bdev, block, size);
 	if (likely(bh) && !buffer_uptodate(bh))
 		bh = __bread_slow(bh);
 	return bh;
 }
 EXPORT_SYMBOL(__bread);
 /*
  * invalidate_bh_lrus() is called rarely - but not only at unmount.
  * This doesn't race because it runs in each cpu either in irq
  * or with preempt disabled.
  */
 static void invalidate_bh_lru(void *arg)
 {
 	struct bh_lru *b = &get_cpu_var(bh_lrus);
 	int i;
 	for (i = 0; i < BH_LRU_SIZE; i++) {
 		brelse(b->bhs[i]);
 		b->bhs[i] = NULL;
 	}
 	put_cpu_var(bh_lrus);
 }
 void invalidate_bh_lrus(void)
 {
 	on_each_cpu(invalidate_bh_lru, NULL, 1);
 }
 EXPORT_SYMBOL_GPL(invalidate_bh_lrus);
 void set_bh_page(struct buffer_head *bh,
 		struct page *page, unsigned long offset)
 {
 	bh->b_page = page;
 	BUG_ON(offset >= PAGE_SIZE);
 	if (PageHighMem(page))
 		/*
 		 * This catches illegal uses and preserves the offset:
 		 */
 		bh->b_data = (char *)(0 + offset);
 	else
 		bh->b_data = page_address(page) + offset;
 }
 EXPORT_SYMBOL(set_bh_page);
 /*
  * Called when truncating a buffer on a page completely.
  */
 static void discard_buffer(struct buffer_head * bh)
 {
 	lock_buffer(bh);
 	clear_buffer_dirty(bh);
 	bh->b_bdev = NULL;
 	clear_buffer_mapped(bh);
 	clear_buffer_req(bh);
 	clear_buffer_new(bh);
 	clear_buffer_delay(bh);
 	clear_buffer_unwritten(bh);
 	unlock_buffer(bh);
 }
 /**
  * block_invalidatepage - invalidate part of all of a buffer-backed page
  *
  * @page: the page which is affected
  * @offset: the index of the truncation point
  *
  * block_invalidatepage() is called when all or part of the page has become
  * invalidatedby a truncate operation.
  *
  * block_invalidatepage() does not have to release all buffers, but it must
  * ensure that no dirty buffer is left outside @offset and that no I/O
  * is underway against any of the blocks which are outside the truncation
  * point.  Because the caller is about to free (and possibly reuse) those
  * blocks on-disk.
  */
 void block_invalidatepage(struct page *page, unsigned long offset)
 {
 	struct buffer_head *head, *bh, *next;
 	unsigned int curr_off = 0;
 	BUG_ON(!PageLocked(page));
 	if (!page_has_buffers(page))
 		goto out;
 	head = page_buffers(page);
 	bh = head;
 	do {
 		unsigned int next_off = curr_off + bh->b_size;
 		next = bh->b_this_page;
 		/*
 		 * is this block fully invalidated?
 		 */
 		if (offset <= curr_off)
 			discard_buffer(bh);
 		curr_off = next_off;
 		bh = next;
 	} while (bh != head);
 	/*
 	 * We release buffers only if the entire page is being invalidated.
 	 * The get_block cached value has been unconditionally invalidated,
 	 * so real IO is not possible anymore.
 	 */
 	if (offset == 0)
 		try_to_release_page(page, 0);
 out:
 	return;
 }
 EXPORT_SYMBOL(block_invalidatepage);
 /*
  * We attach and possibly dirty the buffers atomically wrt
  * __set_page_dirty_buffers() via private_lock.  try_to_free_buffers
  * is already excluded via the page lock.
  */
 void create_empty_buffers(struct page *page,
 			unsigned long blocksize, unsigned long b_state)
 {
 	struct buffer_head *bh, *head, *tail;
 	head = alloc_page_buffers(page, blocksize, 1);
 	bh = head;
 	do {
 		bh->b_state |= b_state;
 		tail = bh;
 		bh = bh->b_this_page;
 	} while (bh);
 	tail->b_this_page = head;
 	spin_lock(&page->mapping->private_lock);
 	if (PageUptodate(page) || PageDirty(page)) {
 		bh = head;
 		do {
 			if (PageDirty(page))
 				set_buffer_dirty(bh);
 			if (PageUptodate(page))
 				set_buffer_uptodate(bh);
 			bh = bh->b_this_page;
 		} while (bh != head);
 	}
 	attach_page_buffers(page, head);
 	spin_unlock(&page->mapping->private_lock);
 }
 EXPORT_SYMBOL(create_empty_buffers);
 /*
  * We are taking a block for data and we don't want any output from any
  * buffer-cache aliases starting from return from that function and
  * until the moment when something will explicitly mark the buffer
  * dirty (hopefully that will not happen until we will free that block ;-)
  * We don't even need to mark it not-uptodate - nobody can expect
  * anything from a newly allocated buffer anyway. We used to used
  * unmap_buffer() for such invalidation, but that was wrong. We definitely
  * don't want to mark the alias unmapped, for example - it would confuse
  * anyone who might pick it with bread() afterwards...
  *
  * Also..  Note that bforget() doesn't lock the buffer.  So there can
  * be writeout I/O going on against recently-freed buffers.  We don't
  * wait on that I/O in bforget() - it's more efficient to wait on the I/O
  * only if we really need to.  That happens here.
  */
 void unmap_underlying_metadata(struct block_device *bdev, sector_t block)
 {
 	struct buffer_head *old_bh;
 	might_sleep();
 	old_bh = __find_get_block_slow(bdev, block);
 	if (old_bh) {
 		clear_buffer_dirty(old_bh);
 		wait_on_buffer(old_bh);
 		clear_buffer_req(old_bh);
 		__brelse(old_bh);
 	}
 }
 EXPORT_SYMBOL(unmap_underlying_metadata);
 /*
  * NOTE! All mapped/uptodate combinations are valid:
  *
  *	Mapped	Uptodate	Meaning
  *
  *	No	No		"unknown" - must do get_block()
  *	No	Yes		"hole" - zero-filled
  *	Yes	No		"allocated" - allocated on disk, not read in
  *	Yes	Yes		"valid" - allocated and up-to-date in memory.
  *
  * "Dirty" is valid only with the last case (mapped+uptodate).
  */
 /*
  * While block_write_full_page is writing back the dirty buffers under
  * the page lock, whoever dirtied the buffers may decide to clean them
  * again at any time.  We handle that by only looking at the buffer
  * state inside lock_buffer().
  *
  * If block_write_full_page() is called for regular writeback
  * (wbc->sync_mode == WB_SYNC_NONE) then it will redirty a page which has a
  * locked buffer.   This only can happen if someone has written the buffer
  * directly, with submit_bh().  At the address_space level PageWriteback
  * prevents this contention from occurring.
  *
  * If block_write_full_page() is called with wbc->sync_mode ==
  * WB_SYNC_ALL, the writes are posted using WRITE_SYNC_PLUG; this
  * causes the writes to be flagged as synchronous writes, but the
  * block device queue will NOT be unplugged, since usually many pages
  * will be pushed to the out before the higher-level caller actually
  * waits for the writes to be completed.  The various wait functions,
  * such as wait_on_writeback_range() will ultimately call sync_page()
  * which will ultimately call blk_run_backing_dev(), which will end up
  * unplugging the device queue.
  */
 static int __block_write_full_page(struct inode *inode, struct page *page,
 			get_block_t *get_block, struct writeback_control *wbc,
 			bh_end_io_t *handler)
 {
 	int err;
 	sector_t block;
 	sector_t last_block;
 	struct buffer_head *bh, *head;
 	const unsigned blocksize = 1 << inode->i_blkbits;
 	int nr_underway = 0;
 	int write_op = (wbc->sync_mode == WB_SYNC_ALL ?
 			WRITE_SYNC_PLUG : WRITE);
 	BUG_ON(!PageLocked(page));
 	last_block = (i_size_read(inode) - 1) >> inode->i_blkbits;
 	if (!page_has_buffers(page)) {
 		create_empty_buffers(page, blocksize,
 					(1 << BH_Dirty)|(1 << BH_Uptodate));
 	}
 	/*
 	 * Be very careful.  We have no exclusion from __set_page_dirty_buffers
 	 * here, and the (potentially unmapped) buffers may become dirty at
 	 * any time.  If a buffer becomes dirty here after we've inspected it
 	 * then we just miss that fact, and the page stays dirty.
 	 *
 	 * Buffers outside i_size may be dirtied by __set_page_dirty_buffers;
 	 * handle that here by just cleaning them.
 	 */
 	block = (sector_t)page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
 	head = page_buffers(page);
 	bh = head;
 	/*
 	 * Get all the dirty buffers mapped to disk addresses and
 	 * handle any aliases from the underlying blockdev's mapping.
 	 */
 	do {
 		if (block > last_block) {
 			/*
 			 * mapped buffers outside i_size will occur, because
 			 * this page can be outside i_size when there is a
 			 * truncate in progress.
 			 */
 			/*
 			 * The buffer was zeroed by block_write_full_page()
 			 */
 			clear_buffer_dirty(bh);
 			set_buffer_uptodate(bh);
 		} else if ((!buffer_mapped(bh) || buffer_delay(bh)) &&
 			   buffer_dirty(bh)) {
 			WARN_ON(bh->b_size != blocksize);
 			err = get_block(inode, block, bh, 1);
 			if (err)
 				goto recover;
 			clear_buffer_delay(bh);
 			if (buffer_new(bh)) {
 				/* blockdev mappings never come here */
 				clear_buffer_new(bh);
 				unmap_underlying_metadata(bh->b_bdev,
 							bh->b_blocknr);
 			}
 		}
 		bh = bh->b_this_page;
 		block++;
 	} while (bh != head);
 	do {
 		if (!buffer_mapped(bh))
 			continue;
 		/*
 		 * If it's a fully non-blocking write attempt and we cannot
 		 * lock the buffer then redirty the page.  Note that this can
 		 * potentially cause a busy-wait loop from writeback threads
 		 * and kswapd activity, but those code paths have their own
 		 * higher-level throttling.
 		 */
 		if (wbc->sync_mode != WB_SYNC_NONE || !wbc->nonblocking) {
 			lock_buffer(bh);
 		} else if (!trylock_buffer(bh)) {
 			redirty_page_for_writepage(wbc, page);
 			continue;
 		}
 		if (test_clear_buffer_dirty(bh)) {
 			mark_buffer_async_write_endio(bh, handler);
 		} else {
 			unlock_buffer(bh);
 		}
 	} while ((bh = bh->b_this_page) != head);
 	/*
 	 * The page and its buffers are protected by PageWriteback(), so we can
 	 * drop the bh refcounts early.
 	 */
 	BUG_ON(PageWriteback(page));
 	set_page_writeback(page);
 	do {
 		struct buffer_head *next = bh->b_this_page;
 		if (buffer_async_write(bh)) {
 			submit_bh(write_op, bh);
 			nr_underway++;
 		}
 		bh = next;
 	} while (bh != head);
 	unlock_page(page);
 	err = 0;
 done:
 	if (nr_underway == 0) {
 		/*
 		 * The page was marked dirty, but the buffers were
 		 * clean.  Someone wrote them back by hand with
 		 * ll_rw_block/submit_bh.  A rare case.
 		 */
 		end_page_writeback(page);
 		/*
 		 * The page and buffer_heads can be released at any time from
 		 * here on.
 		 */
 	}
 	return err;
 recover:
 	/*
 	 * ENOSPC, or some other error.  We may already have added some
 	 * blocks to the file, so we need to write these out to avoid
 	 * exposing stale data.
 	 * The page is currently locked and not marked for writeback
 	 */
 	bh = head;
 	/* Recovery: lock and submit the mapped buffers */
 	do {
 		if (buffer_mapped(bh) && buffer_dirty(bh) &&
 		    !buffer_delay(bh)) {
 			lock_buffer(bh);
 			mark_buffer_async_write_endio(bh, handler);
 		} else {
 			/*
 			 * The buffer may have been set dirty during
 			 * attachment to a dirty page.
 			 */
 			clear_buffer_dirty(bh);
 		}
 	} while ((bh = bh->b_this_page) != head);
 	SetPageError(page);
 	BUG_ON(PageWriteback(page));
 	mapping_set_error(page->mapping, err);
 	set_page_writeback(page);
 	do {
 		struct buffer_head *next = bh->b_this_page;
 		if (buffer_async_write(bh)) {
 			clear_buffer_dirty(bh);
 			submit_bh(write_op, bh);
 			nr_underway++;
 		}
 		bh = next;
 	} while (bh != head);
 	unlock_page(page);
 	goto done;
 }
 /*
  * If a page has any new buffers, zero them out here, and mark them uptodate
  * and dirty so they'll be written out (in order to prevent uninitialised
  * block data from leaking). And clear the new bit.
  */
 void page_zero_new_buffers(struct page *page, unsigned from, unsigned to)
 {
 	unsigned int block_start, block_end;
 	struct buffer_head *head, *bh;
 	BUG_ON(!PageLocked(page));
 	if (!page_has_buffers(page))
 		return;
 	bh = head = page_buffers(page);
 	block_start = 0;
 	do {
 		block_end = block_start + bh->b_size;
 		if (buffer_new(bh)) {
 			if (block_end > from && block_start < to) {
 				if (!PageUptodate(page)) {
 					unsigned start, size;
 					start = max(from, block_start);
 					size = min(to, block_end) - start;
 					zero_user(page, start, size);
 					set_buffer_uptodate(bh);
 				}
 				clear_buffer_new(bh);
 				mark_buffer_dirty(bh);
 			}
 		}
 		block_start = block_end;
 		bh = bh->b_this_page;
 	} while (bh != head);
 }
 EXPORT_SYMBOL(page_zero_new_buffers);
 int block_prepare_write(struct page *page, unsigned from, unsigned to,
 		get_block_t *get_block)
 {
 	struct inode *inode = page->mapping->host;
 	unsigned block_start, block_end;
 	sector_t block;
 	int err = 0;
 	unsigned blocksize, bbits;
 	struct buffer_head *bh, *head, *wait[2], **wait_bh=wait;
 	BUG_ON(!PageLocked(page));
 	BUG_ON(from > PAGE_CACHE_SIZE);
 	BUG_ON(to > PAGE_CACHE_SIZE);
 	BUG_ON(from > to);
 	blocksize = 1 << inode->i_blkbits;
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, blocksize, 0);
 	head = page_buffers(page);
 	bbits = inode->i_blkbits;
 	block = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits);
 	for(bh = head, block_start = 0; bh != head || !block_start;
 	    block++, block_start=block_end, bh = bh->b_this_page) {
 		block_end = block_start + blocksize;
 		if (block_end <= from || block_start >= to) {
 			if (PageUptodate(page)) {
 				if (!buffer_uptodate(bh))
 					set_buffer_uptodate(bh);
 			}
 			continue;
 		}
 		if (buffer_new(bh))
 			clear_buffer_new(bh);
 		if (!buffer_mapped(bh)) {
 			WARN_ON(bh->b_size != blocksize);
 			err = get_block(inode, block, bh, 1);
 			if (err)
 				break;
 			if (buffer_new(bh)) {
 				unmap_underlying_metadata(bh->b_bdev,
 							bh->b_blocknr);
 				if (PageUptodate(page)) {
 					clear_buffer_new(bh);
 					set_buffer_uptodate(bh);
 					mark_buffer_dirty(bh);
 					continue;
 				}
 				if (block_end > to || block_start < from)
 					zero_user_segments(page,
 						to, block_end,
 						block_start, from);
 				continue;
 			}
 		}
 		if (PageUptodate(page)) {
 			if (!buffer_uptodate(bh))
 				set_buffer_uptodate(bh);
 			continue;
 		}
 		if (!buffer_uptodate(bh) && !buffer_delay(bh) &&
 		    !buffer_unwritten(bh) &&
 		     (block_start < from || block_end > to)) {
 			ll_rw_block(READ, 1, &bh);
 			*wait_bh++=bh;
 		}
 	}
 	/*
 	 * If we issued read requests - let them complete.
 	 */
 	while(wait_bh > wait) {
 		wait_on_buffer(*--wait_bh);
 		if (!buffer_uptodate(*wait_bh))
 			err = -EIO;
 	}
 	if (unlikely(err)) {
 		page_zero_new_buffers(page, from, to);
 		ClearPageUptodate(page);
 	}
 	return err;
 }
 EXPORT_SYMBOL(block_prepare_write);
 static int __block_commit_write(struct inode *inode, struct page *page,
 		unsigned from, unsigned to)
 {
 	unsigned block_start, block_end;
 	int partial = 0;
 	unsigned blocksize;
 	struct buffer_head *bh, *head;
 	blocksize = 1 << inode->i_blkbits;
 	for(bh = head = page_buffers(page), block_start = 0;
 	    bh != head || !block_start;
 	    block_start=block_end, bh = bh->b_this_page) {
 		block_end = block_start + blocksize;
 		if (block_end <= from || block_start >= to) {
 			if (!buffer_uptodate(bh))
 				partial = 1;
 		} else {
 			set_buffer_uptodate(bh);
 			mark_buffer_dirty(bh);
 		}
 		clear_buffer_new(bh);
 	}
 	/*
 	 * If this is a partial write which happened to make all buffers
 	 * uptodate then we can optimize away a bogus readpage() for
 	 * the next read(). Here we 'discover' whether the page went
 	 * uptodate as a result of this (potentially partial) write.
 	 */
 	if (!partial)
 		SetPageUptodate(page);
 	return 0;
 }
 int __block_write_begin(struct page *page, loff_t pos, unsigned len,
 		get_block_t *get_block)
 {
 	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
 	return block_prepare_write(page, start, start + len, get_block);
 }
 EXPORT_SYMBOL(__block_write_begin);
 /*
  * block_write_begin takes care of the basic task of block allocation and
  * bringing partial write blocks uptodate first.
  *
  * The filesystem needs to handle block truncation upon failure.
  */
 int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
 		unsigned flags, struct page **pagep, get_block_t *get_block)
 {
 	pgoff_t index = pos >> PAGE_CACHE_SHIFT;
 	struct page *page;
 	int status;
 	page = grab_cache_page_write_begin(mapping, index, flags);
 	if (!page)
 		return -ENOMEM;
 	status = __block_write_begin(page, pos, len, get_block);
 	if (unlikely(status)) {
 		unlock_page(page);
 		page_cache_release(page);
 		page = NULL;
 	}
 	*pagep = page;
 	return status;
 }
 EXPORT_SYMBOL(block_write_begin);
 int block_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
 			struct page *page, void *fsdata)
 {
 	struct inode *inode = mapping->host;
 	unsigned start;
 	start = pos & (PAGE_CACHE_SIZE - 1);
 	if (unlikely(copied < len)) {
 		/*
 		 * The buffers that were written will now be uptodate, so we
 		 * don't have to worry about a readpage reading them and
 		 * overwriting a partial write. However if we have encountered
 		 * a short write and only partially written into a buffer, it
 		 * will not be marked uptodate, so a readpage might come in and
 		 * destroy our partial write.
 		 *
 		 * Do the simplest thing, and just treat any short write to a
 		 * non uptodate page as a zero-length write, and force the
 		 * caller to redo the whole thing.
 		 */
 		if (!PageUptodate(page))
 			copied = 0;
 		page_zero_new_buffers(page, start+copied, start+len);
 	}
 	flush_dcache_page(page);
 	/* This could be a short (even 0-length) commit */
 	__block_commit_write(inode, page, start, start+copied);
 	return copied;
 }
 EXPORT_SYMBOL(block_write_end);
 int generic_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
 			struct page *page, void *fsdata)
 {
 	struct inode *inode = mapping->host;
 	int i_size_changed = 0;
 	copied = block_write_end(file, mapping, pos, len, copied, page, fsdata);
 	/*
 	 * No need to use i_size_read() here, the i_size
 	 * cannot change under us because we hold i_mutex.
 	 *
 	 * But it's important to update i_size while still holding page lock:
 	 * page writeout could otherwise come in and zero beyond i_size.
 	 */
 	if (pos+copied > inode->i_size) {
 		i_size_write(inode, pos+copied);
 		i_size_changed = 1;
 	}
 	unlock_page(page);
 	page_cache_release(page);
 	/*
 	 * Don't mark the inode dirty under page lock. First, it unnecessarily
 	 * makes the holding time of page lock longer. Second, it forces lock
 	 * ordering of page lock and transaction start for journaling
 	 * filesystems.
 	 */
 	if (i_size_changed)
 		mark_inode_dirty(inode);
 	return copied;
 }
 EXPORT_SYMBOL(generic_write_end);
 /*
  * block_is_partially_uptodate checks whether buffers within a page are
  * uptodate or not.
  *
  * Returns true if all buffers which correspond to a file portion
  * we want to read are uptodate.
  */
 int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc,
 					unsigned long from)
 {
 	struct inode *inode = page->mapping->host;
 	unsigned block_start, block_end, blocksize;
 	unsigned to;
 	struct buffer_head *bh, *head;
 	int ret = 1;
 	if (!page_has_buffers(page))
 		return 0;
 	blocksize = 1 << inode->i_blkbits;
 	to = min_t(unsigned, PAGE_CACHE_SIZE - from, desc->count);
 	to = from + to;
 	if (from < blocksize && to > PAGE_CACHE_SIZE - blocksize)
 		return 0;
 	head = page_buffers(page);
 	bh = head;
 	block_start = 0;
 	do {
 		block_end = block_start + blocksize;
 		if (block_end > from && block_start < to) {
 			if (!buffer_uptodate(bh)) {
 				ret = 0;
 				break;
 			}
 			if (block_end >= to)
 				break;
 		}
 		block_start = block_end;
 		bh = bh->b_this_page;
 	} while (bh != head);
 	return ret;
 }
 EXPORT_SYMBOL(block_is_partially_uptodate);
 /*
  * Generic "read page" function for block devices that have the normal
  * get_block functionality. This is most of the block device filesystems.
  * Reads the page asynchronously --- the unlock_buffer() and
  * set/clear_buffer_uptodate() functions propagate buffer state into the
  * page struct once IO has completed.
  */
 int block_read_full_page(struct page *page, get_block_t *get_block)
 {
 	struct inode *inode = page->mapping->host;
 	sector_t iblock, lblock;
 	struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE];
 	unsigned int blocksize;
 	int nr, i;
 	int fully_mapped = 1;
 	BUG_ON(!PageLocked(page));
 	blocksize = 1 << inode->i_blkbits;
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, blocksize, 0);
 	head = page_buffers(page);
 	iblock = (sector_t)page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
 	lblock = (i_size_read(inode)+blocksize-1) >> inode->i_blkbits;
 	bh = head;
 	nr = 0;
 	i = 0;
 	do {
 		if (buffer_uptodate(bh))
 			continue;
 		if (!buffer_mapped(bh)) {
 			int err = 0;
 			fully_mapped = 0;
 			if (iblock < lblock) {
 				WARN_ON(bh->b_size != blocksize);
 				err = get_block(inode, iblock, bh, 0);
 				if (err)
 					SetPageError(page);
 			}
 			if (!buffer_mapped(bh)) {
 				zero_user(page, i * blocksize, blocksize);
 				if (!err)
 					set_buffer_uptodate(bh);
 				continue;
 			}
 			/*
 			 * get_block() might have updated the buffer
 			 * synchronously
 			 */
 			if (buffer_uptodate(bh))
 				continue;
 		}
 		arr[nr++] = bh;
 	} while (i++, iblock++, (bh = bh->b_this_page) != head);
 	if (fully_mapped)
 		SetPageMappedToDisk(page);
 	if (!nr) {
 		/*
 		 * All buffers are uptodate - we can set the page uptodate
 		 * as well. But not if get_block() returned an error.
 		 */
 		if (!PageError(page))
 			SetPageUptodate(page);
 		unlock_page(page);
 		return 0;
 	}
 	/* Stage two: lock the buffers */
 	for (i = 0; i < nr; i++) {
 		bh = arr[i];
 		lock_buffer(bh);
 		mark_buffer_async_read(bh);
 	}
 	/*
 	 * Stage 3: start the IO.  Check for uptodateness
 	 * inside the buffer lock in case another process reading
 	 * the underlying blockdev brought it uptodate (the sct fix).
 	 */
 	for (i = 0; i < nr; i++) {
 		bh = arr[i];
 		if (buffer_uptodate(bh))
 			end_buffer_async_read(bh, 1);
 		else
 			submit_bh(READ, bh);
 	}
 	return 0;
 }
 EXPORT_SYMBOL(block_read_full_page);
 /* utility function for filesystems that need to do work on expanding
  * truncates.  Uses filesystem pagecache writes to allow the filesystem to
  * deal with the hole.
  */
 int generic_cont_expand_simple(struct inode *inode, loff_t size)
 {
 	struct address_space *mapping = inode->i_mapping;
 	struct page *page;
 	void *fsdata;
 	int err;
 	err = inode_newsize_ok(inode, size);
 	if (err)
 		goto out;
 	err = pagecache_write_begin(NULL, mapping, size, 0,
 				AOP_FLAG_UNINTERRUPTIBLE|AOP_FLAG_CONT_EXPAND,
 				&page, &fsdata);
 	if (err)
 		goto out;
 	err = pagecache_write_end(NULL, mapping, size, 0, 0, page, fsdata);
 	BUG_ON(err > 0);
 out:
 	return err;
 }
 EXPORT_SYMBOL(generic_cont_expand_simple);
 static int cont_expand_zero(struct file *file, struct address_space *mapping,
 			    loff_t pos, loff_t *bytes)
 {
 	struct inode *inode = mapping->host;
 	unsigned blocksize = 1 << inode->i_blkbits;
 	struct page *page;
 	void *fsdata;
 	pgoff_t index, curidx;
 	loff_t curpos;
 	unsigned zerofrom, offset, len;
 	int err = 0;
 	index = pos >> PAGE_CACHE_SHIFT;
 	offset = pos & ~PAGE_CACHE_MASK;
 	while (index > (curidx = (curpos = *bytes)>>PAGE_CACHE_SHIFT)) {
 		zerofrom = curpos & ~PAGE_CACHE_MASK;
 		if (zerofrom & (blocksize-1)) {
 			*bytes |= (blocksize-1);
 			(*bytes)++;
 		}
 		len = PAGE_CACHE_SIZE - zerofrom;
 		err = pagecache_write_begin(file, mapping, curpos, len,
 						AOP_FLAG_UNINTERRUPTIBLE,
 						&page, &fsdata);
 		if (err)
 			goto out;
 		zero_user(page, zerofrom, len);
 		err = pagecache_write_end(file, mapping, curpos, len, len,
 						page, fsdata);
 		if (err < 0)
 			goto out;
 		BUG_ON(err != len);
 		err = 0;
 		balance_dirty_pages_ratelimited(mapping);
 	}
 	/* page covers the boundary, find the boundary offset */
 	if (index == curidx) {
 		zerofrom = curpos & ~PAGE_CACHE_MASK;
 		/* if we will expand the thing last block will be filled */
 		if (offset <= zerofrom) {
 			goto out;
 		}
 		if (zerofrom & (blocksize-1)) {
 			*bytes |= (blocksize-1);
 			(*bytes)++;
 		}
 		len = offset - zerofrom;
 		err = pagecache_write_begin(file, mapping, curpos, len,
 						AOP_FLAG_UNINTERRUPTIBLE,
 						&page, &fsdata);
 		if (err)
 			goto out;
 		zero_user(page, zerofrom, len);
 		err = pagecache_write_end(file, mapping, curpos, len, len,
 						page, fsdata);
 		if (err < 0)
 			goto out;
 		BUG_ON(err != len);
 		err = 0;
 	}
 out:
 	return err;
 }
 /*
  * For moronic filesystems that do not allow holes in file.
  * We may have to extend the file.
  */
 int cont_write_begin(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned flags,
 			struct page **pagep, void **fsdata,
 			get_block_t *get_block, loff_t *bytes)
 {
 	struct inode *inode = mapping->host;
 	unsigned blocksize = 1 << inode->i_blkbits;
 	unsigned zerofrom;
 	int err;
 	err = cont_expand_zero(file, mapping, pos, bytes);
 	if (err)
 		return err;
 	zerofrom = *bytes & ~PAGE_CACHE_MASK;
 	if (pos+len > *bytes && zerofrom & (blocksize-1)) {
 		*bytes |= (blocksize-1);
 		(*bytes)++;
 	}
 	return block_write_begin(mapping, pos, len, flags, pagep, get_block);
 }
 EXPORT_SYMBOL(cont_write_begin);
 int block_commit_write(struct page *page, unsigned from, unsigned to)
 {
 	struct inode *inode = page->mapping->host;
 	__block_commit_write(inode,page,from,to);
 	return 0;
 }
 EXPORT_SYMBOL(block_commit_write);
 /*
  * block_page_mkwrite() is not allowed to change the file size as it gets
  * called from a page fault handler when a page is first dirtied. Hence we must
  * be careful to check for EOF conditions here. We set the page up correctly
  * for a written page which means we get ENOSPC checking when writing into
  * holes and correct delalloc and unwritten extent mapping on filesystems that
  * support these features.
  *
  * We are not allowed to take the i_mutex here so we have to play games to
  * protect against truncate races as the page could now be beyond EOF.  Because
  * truncate writes the inode size before removing pages, once we have the
  * page lock we can determine safely if the page is beyond EOF. If it is not
  * beyond EOF, then the page is guaranteed safe against truncation until we
  * unlock the page.
  */
 int
 block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
 		   get_block_t get_block)
 {
 	struct page *page = vmf->page;
 	struct inode *inode = vma->vm_file->f_path.dentry->d_inode;
 	unsigned long end;
 	loff_t size;
 	int ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */
 	lock_page(page);
 	size = i_size_read(inode);
 	if ((page->mapping != inode->i_mapping) ||
 	    (page_offset(page) > size)) {
 		/* page got truncated out from underneath us */
 		unlock_page(page);
 		goto out;
 	}
 	/* page is wholly or partially inside EOF */
 	if (((page->index + 1) << PAGE_CACHE_SHIFT) > size)
 		end = size & ~PAGE_CACHE_MASK;
 	else
 		end = PAGE_CACHE_SIZE;
 	ret = block_prepare_write(page, 0, end, get_block);
 	if (!ret)
 		ret = block_commit_write(page, 0, end);
 	if (unlikely(ret)) {
 		unlock_page(page);
 		if (ret == -ENOMEM)
 			ret = VM_FAULT_OOM;
 		else /* -ENOSPC, -EIO, etc */
 			ret = VM_FAULT_SIGBUS;
 	} else
 		ret = VM_FAULT_LOCKED;
 out:
 	return ret;
 }
 EXPORT_SYMBOL(block_page_mkwrite);
 /*
  * nobh_write_begin()'s prereads are special: the buffer_heads are freed
  * immediately, while under the page lock.  So it needs a special end_io
  * handler which does not touch the bh after unlocking it.
  */
 static void end_buffer_read_nobh(struct buffer_head *bh, int uptodate)
 {
 	__end_buffer_read_notouch(bh, uptodate);
 }
 /*
  * Attach the singly-linked list of buffers created by nobh_write_begin, to
  * the page (converting it to circular linked list and taking care of page
  * dirty races).
  */
 static void attach_nobh_buffers(struct page *page, struct buffer_head *head)
 {
 	struct buffer_head *bh;
 	BUG_ON(!PageLocked(page));
 	spin_lock(&page->mapping->private_lock);
 	bh = head;
 	do {
 		if (PageDirty(page))
 			set_buffer_dirty(bh);
 		if (!bh->b_this_page)
 			bh->b_this_page = head;
 		bh = bh->b_this_page;
 	} while (bh != head);
 	attach_page_buffers(page, head);
 	spin_unlock(&page->mapping->private_lock);
 }
 /*
  * On entry, the page is fully not uptodate.
  * On exit the page is fully uptodate in the areas outside (from,to)
  * The filesystem needs to handle block truncation upon failure.
  */
 int nobh_write_begin(struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned flags,
 			struct page **pagep, void **fsdata,
 			get_block_t *get_block)
 {
 	struct inode *inode = mapping->host;
 	const unsigned blkbits = inode->i_blkbits;
 	const unsigned blocksize = 1 << blkbits;
 	struct buffer_head *head, *bh;
 	struct page *page;
 	pgoff_t index;
 	unsigned from, to;
 	unsigned block_in_page;
 	unsigned block_start, block_end;
 	sector_t block_in_file;
 	int nr_reads = 0;
 	int ret = 0;
 	int is_mapped_to_disk = 1;
 	index = pos >> PAGE_CACHE_SHIFT;
 	from = pos & (PAGE_CACHE_SIZE - 1);
 	to = from + len;
 	page = grab_cache_page_write_begin(mapping, index, flags);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
 	*fsdata = NULL;
 	if (page_has_buffers(page)) {
 		unlock_page(page);
 		page_cache_release(page);
 		*pagep = NULL;
 		return block_write_begin(mapping, pos, len, flags, pagep,
 					 get_block);
 	}
 	if (PageMappedToDisk(page))
 		return 0;
 	/*
 	 * Allocate buffers so that we can keep track of state, and potentially
 	 * attach them to the page if an error occurs. In the common case of
 	 * no error, they will just be freed again without ever being attached
 	 * to the page (which is all OK, because we're under the page lock).
 	 *
 	 * Be careful: the buffer linked list is a NULL terminated one, rather
 	 * than the circular one we're used to.
 	 */
 	head = alloc_page_buffers(page, blocksize, 0);
 	if (!head) {
 		ret = -ENOMEM;
 		goto out_release;
 	}
 	block_in_file = (sector_t)page->index << (PAGE_CACHE_SHIFT - blkbits);
 	/*
 	 * We loop across all blocks in the page, whether or not they are
 	 * part of the affected region.  This is so we can discover if the
 	 * page is fully mapped-to-disk.
 	 */
 	for (block_start = 0, block_in_page = 0, bh = head;
 		  block_start < PAGE_CACHE_SIZE;
 		  block_in_page++, block_start += blocksize, bh = bh->b_this_page) {
 		int create;
 		block_end = block_start + blocksize;
 		bh->b_state = 0;
 		create = 1;
 		if (block_start >= to)
 			create = 0;
 		ret = get_block(inode, block_in_file + block_in_page,
 					bh, create);
 		if (ret)
 			goto failed;
 		if (!buffer_mapped(bh))
 			is_mapped_to_disk = 0;
 		if (buffer_new(bh))
 			unmap_underlying_metadata(bh->b_bdev, bh->b_blocknr);
 		if (PageUptodate(page)) {
 			set_buffer_uptodate(bh);
 			continue;
 		}
 		if (buffer_new(bh) || !buffer_mapped(bh)) {
 			zero_user_segments(page, block_start, from,
 							to, block_end);
 			continue;
 		}
 		if (buffer_uptodate(bh))
 			continue;	/* reiserfs does this */
 		if (block_start < from || block_end > to) {
 			lock_buffer(bh);
 			bh->b_end_io = end_buffer_read_nobh;
 			submit_bh(READ, bh);
 			nr_reads++;
 		}
 	}
 	if (nr_reads) {
 		/*
 		 * The page is locked, so these buffers are protected from
 		 * any VM or truncate activity.  Hence we don't need to care
 		 * for the buffer_head refcounts.
 		 */
 		for (bh = head; bh; bh = bh->b_this_page) {
 			wait_on_buffer(bh);
 			if (!buffer_uptodate(bh))
 				ret = -EIO;
 		}
 		if (ret)
 			goto failed;
 	}
 	if (is_mapped_to_disk)
 		SetPageMappedToDisk(page);
 	*fsdata = head; /* to be released by nobh_write_end */
 	return 0;
 failed:
 	BUG_ON(!ret);
 	/*
 	 * Error recovery is a bit difficult. We need to zero out blocks that
 	 * were newly allocated, and dirty them to ensure they get written out.
 	 * Buffers need to be attached to the page at this point, otherwise
 	 * the handling of potential IO errors during writeout would be hard
 	 * (could try doing synchronous writeout, but what if that fails too?)
 	 */
 	attach_nobh_buffers(page, head);
 	page_zero_new_buffers(page, from, to);
 out_release:
 	unlock_page(page);
 	page_cache_release(page);
 	*pagep = NULL;
 	return ret;
 }
 EXPORT_SYMBOL(nobh_write_begin);
 int nobh_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
 			struct page *page, void *fsdata)
 {
 	struct inode *inode = page->mapping->host;
 	struct buffer_head *head = fsdata;
 	struct buffer_head *bh;
 	BUG_ON(fsdata != NULL && page_has_buffers(page));
 	if (unlikely(copied < len) && head)
 		attach_nobh_buffers(page, head);
 	if (page_has_buffers(page))
 		return generic_write_end(file, mapping, pos, len,
 					copied, page, fsdata);
 	SetPageUptodate(page);
 	set_page_dirty(page);
 	if (pos+copied > inode->i_size) {
 		i_size_write(inode, pos+copied);
 		mark_inode_dirty(inode);
 	}
 	unlock_page(page);
 	page_cache_release(page);
 	while (head) {
 		bh = head;
 		head = head->b_this_page;
 		free_buffer_head(bh);
 	}
 	return copied;
 }
 EXPORT_SYMBOL(nobh_write_end);
 /*
  * nobh_writepage() - based on block_full_write_page() except
  * that it tries to operate without attaching bufferheads to
  * the page.
  */
 int nobh_writepage(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc)
 {
 	struct inode * const inode = page->mapping->host;
 	loff_t i_size = i_size_read(inode);
 	const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 	unsigned offset;
 	int ret;
 	/* Is the page fully inside i_size? */
 	if (page->index < end_index)
 		goto out;
 	/* Is the page fully outside i_size? (truncate in progress) */
 	offset = i_size & (PAGE_CACHE_SIZE-1);
 	if (page->index >= end_index+1 || !offset) {
 		/*
 		 * The page may have dirty, unmapped buffers.  For example,
 		 * they may have been added in ext3_writepage().  Make them
 		 * freeable here, so the page does not leak.
 		 */
 #if 0
 		/* Not really sure about this  - do we need this ? */
 		if (page->mapping->a_ops->invalidatepage)
 			page->mapping->a_ops->invalidatepage(page, offset);
 #endif
 		unlock_page(page);
 		return 0; /* don't care */
 	}
 	/*
 	 * The page straddles i_size.  It must be zeroed out on each and every
 	 * writepage invocation because it may be mmapped.  "A file is mapped
 	 * in multiples of the page size.  For a file that is not a multiple of
 	 * the  page size, the remaining memory is zeroed when mapped, and
 	 * writes to that region are not written out to the file."
 	 */
 	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
 out:
 	ret = mpage_writepage(page, get_block, wbc);
 	if (ret == -EAGAIN)
 		ret = __block_write_full_page(inode, page, get_block, wbc,
 					      end_buffer_async_write);
 	return ret;
 }
 EXPORT_SYMBOL(nobh_writepage);
 int nobh_truncate_page(struct address_space *mapping,
 			loff_t from, get_block_t *get_block)
 {
 	pgoff_t index = from >> PAGE_CACHE_SHIFT;
 	unsigned offset = from & (PAGE_CACHE_SIZE-1);
 	unsigned blocksize;
 	sector_t iblock;
 	unsigned length, pos;
 	struct inode *inode = mapping->host;
 	struct page *page;
 	struct buffer_head map_bh;
 	int err;
 	blocksize = 1 << inode->i_blkbits;
 	length = offset & (blocksize - 1);
 	/* Block boundary? Nothing to do */
 	if (!length)
 		return 0;
 	length = blocksize - length;
 	iblock = (sector_t)index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
 	page = grab_cache_page(mapping, index);
 	err = -ENOMEM;
 	if (!page)
 		goto out;
 	if (page_has_buffers(page)) {
 has_buffers:
 		unlock_page(page);
 		page_cache_release(page);
 		return block_truncate_page(mapping, from, get_block);
 	}
 	/* Find the buffer that contains "offset" */
 	pos = blocksize;
 	while (offset >= pos) {
 		iblock++;
 		pos += blocksize;
 	}
 	map_bh.b_size = blocksize;
 	map_bh.b_state = 0;
 	err = get_block(inode, iblock, &map_bh, 0);
 	if (err)
 		goto unlock;
 	/* unmapped? It's a hole - nothing to do */
 	if (!buffer_mapped(&map_bh))
 		goto unlock;
 	/* Ok, it's mapped. Make sure it's up-to-date */
 	if (!PageUptodate(page)) {
 		err = mapping->a_ops->readpage(NULL, page);
 		if (err) {
 			page_cache_release(page);
 			goto out;
 		}
 		lock_page(page);
 		if (!PageUptodate(page)) {
 			err = -EIO;
 			goto unlock;
 		}
 		if (page_has_buffers(page))
 			goto has_buffers;
 	}
 	zero_user(page, offset, length);
 	set_page_dirty(page);
 	err = 0;
 unlock:
 	unlock_page(page);
 	page_cache_release(page);
 out:
 	return err;
 }
 EXPORT_SYMBOL(nobh_truncate_page);
 int block_truncate_page(struct address_space *mapping,
 			loff_t from, get_block_t *get_block)
 {
 	pgoff_t index = from >> PAGE_CACHE_SHIFT;
 	unsigned offset = from & (PAGE_CACHE_SIZE-1);
 	unsigned blocksize;
 	sector_t iblock;
 	unsigned length, pos;
 	struct inode *inode = mapping->host;
 	struct page *page;
 	struct buffer_head *bh;
 	int err;
 	blocksize = 1 << inode->i_blkbits;
 	length = offset & (blocksize - 1);
 	/* Block boundary? Nothing to do */
 	if (!length)
 		return 0;
 	length = blocksize - length;
 	iblock = (sector_t)index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
 	page = grab_cache_page(mapping, index);
 	err = -ENOMEM;
 	if (!page)
 		goto out;
 	if (!page_has_buffers(page))
 		create_empty_buffers(page, blocksize, 0);
 	/* Find the buffer that contains "offset" */
 	bh = page_buffers(page);
 	pos = blocksize;
 	while (offset >= pos) {
 		bh = bh->b_this_page;
 		iblock++;
 		pos += blocksize;
 	}
 	err = 0;
 	if (!buffer_mapped(bh)) {
 		WARN_ON(bh->b_size != blocksize);
 		err = get_block(inode, iblock, bh, 0);
 		if (err)
 			goto unlock;
 		/* unmapped? It's a hole - nothing to do */
 		if (!buffer_mapped(bh))
 			goto unlock;
 	}
 	/* Ok, it's mapped. Make sure it's up-to-date */
 	if (PageUptodate(page))
 		set_buffer_uptodate(bh);
 	if (!buffer_uptodate(bh) && !buffer_delay(bh) && !buffer_unwritten(bh)) {
 		err = -EIO;
 		ll_rw_block(READ, 1, &bh);
 		wait_on_buffer(bh);
 		/* Uhhuh. Read error. Complain and punt. */
 		if (!buffer_uptodate(bh))
 			goto unlock;
 	}
 	zero_user(page, offset, length);
 	mark_buffer_dirty(bh);
 	err = 0;
 unlock:
 	unlock_page(page);
 	page_cache_release(page);
 out:
 	return err;
 }
 EXPORT_SYMBOL(block_truncate_page);
 /*
  * The generic ->writepage function for buffer-backed address_spaces
  * this form passes in the end_io handler used to finish the IO.
  */
 int block_write_full_page_endio(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc, bh_end_io_t *handler)
 {
 	struct inode * const inode = page->mapping->host;
 	loff_t i_size = i_size_read(inode);
 	const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 	unsigned offset;
 	/* Is the page fully inside i_size? */
 	if (page->index < end_index)
 		return __block_write_full_page(inode, page, get_block, wbc,
 					       handler);
 	/* Is the page fully outside i_size? (truncate in progress) */
 	offset = i_size & (PAGE_CACHE_SIZE-1);
 	if (page->index >= end_index+1 || !offset) {
 		/*
 		 * The page may have dirty, unmapped buffers.  For example,
 		 * they may have been added in ext3_writepage().  Make them
 		 * freeable here, so the page does not leak.
 		 */
 		do_invalidatepage(page, 0);
 		unlock_page(page);
 		return 0; /* don't care */
 	}
 	/*
 	 * The page straddles i_size.  It must be zeroed out on each and every
 	 * writepage invocation because it may be mmapped.  "A file is mapped
 	 * in multiples of the page size.  For a file that is not a multiple of
 	 * the  page size, the remaining memory is zeroed when mapped, and
 	 * writes to that region are not written out to the file."
 	 */
 	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
 	return __block_write_full_page(inode, page, get_block, wbc, handler);
 }
 EXPORT_SYMBOL(block_write_full_page_endio);
 /*
  * The generic ->writepage function for buffer-backed address_spaces
  */
 int block_write_full_page(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc)
 {
 	return block_write_full_page_endio(page, get_block, wbc,
 					   end_buffer_async_write);
 }
 EXPORT_SYMBOL(block_write_full_page);
 sector_t generic_block_bmap(struct address_space *mapping, sector_t block,
 			    get_block_t *get_block)
 {
 	struct buffer_head tmp;
 	struct inode *inode = mapping->host;
 	tmp.b_state = 0;
 	tmp.b_blocknr = 0;
 	tmp.b_size = 1 << inode->i_blkbits;
 	get_block(inode, block, &tmp, 0);
 	return tmp.b_blocknr;
 }
 EXPORT_SYMBOL(generic_block_bmap);
 static void end_bio_bh_io_sync(struct bio *bio, int err)
 {
 	struct buffer_head *bh = bio->bi_private;
 	if (err == -EOPNOTSUPP) {
 		set_bit(BIO_EOPNOTSUPP, &bio->bi_flags);
 		set_bit(BH_Eopnotsupp, &bh->b_state);
 	}
 	if (unlikely (test_bit(BIO_QUIET,&bio->bi_flags)))
 		set_bit(BH_Quiet, &bh->b_state);
 	bh->b_end_io(bh, test_bit(BIO_UPTODATE, &bio->bi_flags));
 	bio_put(bio);
 }
 int submit_bh(int rw, struct buffer_head * bh)
 {
 	struct bio *bio;
 	int ret = 0;
 	BUG_ON(!buffer_locked(bh));
 	BUG_ON(!buffer_mapped(bh));
 	BUG_ON(!bh->b_end_io);
 	BUG_ON(buffer_delay(bh));
 	BUG_ON(buffer_unwritten(bh));
 	/*
-	 * Mask in barrier bit for a write (could be either a WRITE or a
-	 * WRITE_SYNC
-	 */
-	if (buffer_ordered(bh) && (rw & WRITE))
-		rw |= WRITE_BARRIER;
-	/*
 	 * Only clear out a write error when rewriting
 	 */
 	if (test_set_buffer_req(bh) && (rw & WRITE))
 		clear_buffer_write_io_error(bh);
 	/*
 	 * from here on down, it's all bio -- do the initial mapping,
 	 * submit_bio -> generic_make_request may further map this bio around
 	 */
 	bio = bio_alloc(GFP_NOIO, 1);
 	bio->bi_sector = bh->b_blocknr * (bh->b_size >> 9);
 	bio->bi_bdev = bh->b_bdev;
 	bio->bi_io_vec[0].bv_page = bh->b_page;
 	bio->bi_io_vec[0].bv_len = bh->b_size;
 	bio->bi_io_vec[0].bv_offset = bh_offset(bh);
 	bio->bi_vcnt = 1;
 	bio->bi_idx = 0;
 	bio->bi_size = bh->b_size;
 	bio->bi_end_io = end_bio_bh_io_sync;
 	bio->bi_private = bh;
 	bio_get(bio);
 	submit_bio(rw, bio);
 	if (bio_flagged(bio, BIO_EOPNOTSUPP))
 		ret = -EOPNOTSUPP;
 	bio_put(bio);
 	return ret;
 }
 EXPORT_SYMBOL(submit_bh);
 /**
  * ll_rw_block: low-level access to block devices (DEPRECATED)
  * @rw: whether to %READ or %WRITE or %SWRITE or maybe %READA (readahead)
  * @nr: number of &struct buffer_heads in the array
  * @bhs: array of pointers to &struct buffer_head
  *
  * ll_rw_block() takes an array of pointers to &struct buffer_heads, and
  * requests an I/O operation on them, either a %READ or a %WRITE.  The third
  * %SWRITE is like %WRITE only we make sure that the *current* data in buffers
  * are sent to disk. The fourth %READA option is described in the documentation
  * for generic_make_request() which ll_rw_block() calls.
  *
  * This function drops any buffer that it cannot get a lock on (with the
  * BH_Lock state bit) unless SWRITE is required, any buffer that appears to be
  * clean when doing a write request, and any buffer that appears to be
  * up-to-date when doing read request.  Further it marks as clean buffers that
  * are processed for writing (the buffer cache won't assume that they are
  * actually clean until the buffer gets unlocked).
  *
  * ll_rw_block sets b_end_io to simple completion handler that marks
  * the buffer up-to-date (if approriate), unlocks the buffer and wakes
  * any waiters.
  *
  * All of the buffers must be for the same device, and must also be a
  * multiple of the current approved size for the device.
  */
 void ll_rw_block(int rw, int nr, struct buffer_head *bhs[])
 {
 	int i;
 	for (i = 0; i < nr; i++) {
 		struct buffer_head *bh = bhs[i];
 		if (rw == SWRITE || rw == SWRITE_SYNC || rw == SWRITE_SYNC_PLUG)
 			lock_buffer(bh);
 		else if (!trylock_buffer(bh))
 			continue;
 		if (rw == WRITE || rw == SWRITE || rw == SWRITE_SYNC ||
 		    rw == SWRITE_SYNC_PLUG) {
 			if (test_clear_buffer_dirty(bh)) {
 				bh->b_end_io = end_buffer_write_sync;
 				get_bh(bh);
 				if (rw == SWRITE_SYNC)
 					submit_bh(WRITE_SYNC, bh);
 				else
 					submit_bh(WRITE, bh);
 				continue;
 			}
 		} else {
 			if (!buffer_uptodate(bh)) {
 				bh->b_end_io = end_buffer_read_sync;
 				get_bh(bh);
 				submit_bh(rw, bh);
 				continue;
 			}
 		}
 		unlock_buffer(bh);
 	}
 }
 EXPORT_SYMBOL(ll_rw_block);
 /*
  * For a data-integrity writeout, we need to wait upon any in-progress I/O
  * and then start new I/O and then wait upon it.  The caller must have a ref on
  * the buffer_head.
  */
-int sync_dirty_buffer(struct buffer_head *bh)
+int __sync_dirty_buffer(struct buffer_head *bh, int rw)
 {
 	int ret = 0;
 	WARN_ON(atomic_read(&bh->b_count) < 1);
 	lock_buffer(bh);
 	if (test_clear_buffer_dirty(bh)) {
 		get_bh(bh);
 		bh->b_end_io = end_buffer_write_sync;
-		ret = submit_bh(WRITE_SYNC, bh);
+		ret = submit_bh(rw, bh);
 		wait_on_buffer(bh);
 		if (buffer_eopnotsupp(bh)) {
 			clear_buffer_eopnotsupp(bh);
 			ret = -EOPNOTSUPP;
 		}
 		if (!ret && !buffer_uptodate(bh))
 			ret = -EIO;
 	} else {
 		unlock_buffer(bh);
 	}
 	return ret;
+}
+EXPORT_SYMBOL(__sync_dirty_buffer);
+int sync_dirty_buffer(struct buffer_head *bh)
+{
+	return __sync_dirty_buffer(bh, WRITE_SYNC);
 }
 EXPORT_SYMBOL(sync_dirty_buffer);
 /*
  * try_to_free_buffers() checks if all the buffers on this particular page
  * are unused, and releases them if so.
  *
  * Exclusion against try_to_free_buffers may be obtained by either
  * locking the page or by holding its mapping's private_lock.
  *
  * If the page is dirty but all the buffers are clean then we need to
  * be sure to mark the page clean as well.  This is because the page
  * may be against a block device, and a later reattachment of buffers
  * to a dirty page will set *all* buffers dirty.  Which would corrupt
  * filesystem data on the same device.
  *
  * The same applies to regular filesystem pages: if all the buffers are
  * clean then we set the page clean and proceed.  To do that, we require
  * total exclusion from __set_page_dirty_buffers().  That is obtained with
  * private_lock.
  *
  * try_to_free_buffers() is non-blocking.
  */
 static inline int buffer_busy(struct buffer_head *bh)
 {
 	return atomic_read(&bh->b_count) |
 		(bh->b_state & ((1 << BH_Dirty) | (1 << BH_Lock)));
 }
 static int
 drop_buffers(struct page *page, struct buffer_head **buffers_to_free)
 {
 	struct buffer_head *head = page_buffers(page);
 	struct buffer_head *bh;
 	bh = head;
 	do {
 		if (buffer_write_io_error(bh) && page->mapping)
 			set_bit(AS_EIO, &page->mapping->flags);
 		if (buffer_busy(bh))
 			goto failed;
 		bh = bh->b_this_page;
 	} while (bh != head);
 	do {
 		struct buffer_head *next = bh->b_this_page;
 		if (bh->b_assoc_map)
 			__remove_assoc_queue(bh);
 		bh = next;
 	} while (bh != head);
 	*buffers_to_free = head;
 	__clear_page_buffers(page);
 	return 1;
 failed:
 	return 0;
 }
 int try_to_free_buffers(struct page *page)
 {
 	struct address_space * const mapping = page->mapping;
 	struct buffer_head *buffers_to_free = NULL;
 	int ret = 0;
 	BUG_ON(!PageLocked(page));
 	if (PageWriteback(page))
 		return 0;
 	if (mapping == NULL) {		/* can this still happen? */
 		ret = drop_buffers(page, &buffers_to_free);
 		goto out;
 	}
 	spin_lock(&mapping->private_lock);
 	ret = drop_buffers(page, &buffers_to_free);
 	/*
 	 * If the filesystem writes its buffers by hand (eg ext3)
 	 * then we can have clean buffers against a dirty page.  We
 	 * clean the page here; otherwise the VM will never notice
 	 * that the filesystem did any IO at all.
 	 *
 	 * Also, during truncate, discard_buffer will have marked all
 	 * the page's buffers clean.  We discover that here and clean
 	 * the page also.
 	 *
 	 * private_lock must be held over this entire operation in order
 	 * to synchronise against __set_page_dirty_buffers and prevent the
 	 * dirty bit from being lost.
 	 */
 	if (ret)
 		cancel_dirty_page(page, PAGE_CACHE_SIZE);
 	spin_unlock(&mapping->private_lock);
 out:
 	if (buffers_to_free) {
 		struct buffer_head *bh = buffers_to_free;
 		do {
 			struct buffer_head *next = bh->b_this_page;
 			free_buffer_head(bh);
 			bh = next;
 		} while (bh != buffers_to_free);
 	}
 	return ret;
 }
 EXPORT_SYMBOL(try_to_free_buffers);
 void block_sync_page(struct page *page)
 {
 	struct address_space *mapping;
 	smp_mb();
 	mapping = page_mapping(page);
 	if (mapping)
 		blk_run_backing_dev(mapping->backing_dev_info, page);
 }
 EXPORT_SYMBOL(block_sync_page);
 /*
  * There are no bdflush tunables left.  But distributions are
  * still running obsolete flush daemons, so we terminate them here.
  *
  * Use of bdflush() is deprecated and will be removed in a future kernel.
  * The `flush-X' kernel threads fully replace bdflush daemons and this call.
  */
 SYSCALL_DEFINE2(bdflush, int, func, long, data)
 {
 	static int msg_count;
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 	if (msg_count < 5) {
 		msg_count++;
 		printk(KERN_INFO
 			"warning: process `%s' used the obsolete bdflush"
 			" system call\n", current->comm);
 		printk(KERN_INFO "Fix your initscripts?\n");
 	}
 	if (func == 1)
 		do_exit(0);
 	return 0;
 }
 /*
  * Buffer-head allocation
  */
 static struct kmem_cache *bh_cachep;
 /*
  * Once the number of bh's in the machine exceeds this level, we start
  * stripping them in writeback.
  */
 static int max_buffer_heads;
 int buffer_heads_over_limit;
 struct bh_accounting {
 	int nr;			/* Number of live bh's */
 	int ratelimit;		/* Limit cacheline bouncing */
 };
 static DEFINE_PER_CPU(struct bh_accounting, bh_accounting) = {0, 0};
 static void recalc_bh_state(void)
 {
 	int i;
 	int tot = 0;
 	if (__get_cpu_var(bh_accounting).ratelimit++ < 4096)
 		return;
 	__get_cpu_var(bh_accounting).ratelimit = 0;
 	for_each_online_cpu(i)
 		tot += per_cpu(bh_accounting, i).nr;
 	buffer_heads_over_limit = (tot > max_buffer_heads);
 }
 struct buffer_head *alloc_buffer_head(gfp_t gfp_flags)
 {
 	struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags);
 	if (ret) {
 		INIT_LIST_HEAD(&ret->b_assoc_buffers);
 		get_cpu_var(bh_accounting).nr++;
 		recalc_bh_state();
 		put_cpu_var(bh_accounting);
 	}
 	return ret;
 }
 EXPORT_SYMBOL(alloc_buffer_head);
 void free_buffer_head(struct buffer_head *bh)
 {
 	BUG_ON(!list_empty(&bh->b_assoc_buffers));
 	kmem_cache_free(bh_cachep, bh);
 	get_cpu_var(bh_accounting).nr--;
 	recalc_bh_state();
 	put_cpu_var(bh_accounting);
 }
 EXPORT_SYMBOL(free_buffer_head);
 static void buffer_exit_cpu(int cpu)
 {
 	int i;
 	struct bh_lru *b = &per_cpu(bh_lrus, cpu);
 	for (i = 0; i < BH_LRU_SIZE; i++) {
 		brelse(b->bhs[i]);
 		b->bhs[i] = NULL;
 	}
 	get_cpu_var(bh_accounting).nr += per_cpu(bh_accounting, cpu).nr;
 	per_cpu(bh_accounting, cpu).nr = 0;
 	put_cpu_var(bh_accounting);
 }
 static int buffer_cpu_notify(struct notifier_block *self,
 			      unsigned long action, void *hcpu)
 {
 	if (action == CPU_DEAD || action == CPU_DEAD_FROZEN)
 		buffer_exit_cpu((unsigned long)hcpu);
 	return NOTIFY_OK;
 }
 /**
  * bh_uptodate_or_lock - Test whether the buffer is uptodate
  * @bh: struct buffer_head
  *
  * Return true if the buffer is up-to-date and false,
  * with the buffer locked, if not.
  */
 int bh_uptodate_or_lock(struct buffer_head *bh)
 {
 	if (!buffer_uptodate(bh)) {
 		lock_buffer(bh);
 		if (!buffer_uptodate(bh))
 			return 0;
 		unlock_buffer(bh);
 	}
 	return 1;
 }
 EXPORT_SYMBOL(bh_uptodate_or_lock);
 /**
  * bh_submit_read - Submit a locked buffer for reading
  * @bh: struct buffer_head
  *
  * Returns zero on success and -EIO on error.
  */
 int bh_submit_read(struct buffer_head *bh)
 {
 	BUG_ON(!buffer_locked(bh));
 	if (buffer_uptodate(bh)) {
 		unlock_buffer(bh);
 		return 0;
 	}
 	get_bh(bh);
 	bh->b_end_io = end_buffer_read_sync;
 	submit_bh(READ, bh);
 	wait_on_buffer(bh);
 	if (buffer_uptodate(bh))
 		return 0;
 	return -EIO;
 }
 EXPORT_SYMBOL(bh_submit_read);
 void __init buffer_init(void)
 {
 	int nrpages;
 	bh_cachep = kmem_cache_create("buffer_head",
 			sizeof(struct buffer_head), 0,
 				(SLAB_RECLAIM_ACCOUNT|SLAB_PANIC|
 				SLAB_MEM_SPREAD),
 				NULL);
 	/*
 	 * Limit the bh occupancy to 10% of ZONE_NORMAL

fs/jbd/commit.c

Diff comments View file @ 87e9951

 /*
  * linux/fs/jbd/commit.c
  *
  * Written by Stephen C. Tweedie <sct@redhat.com>, 1998
  *
  * Copyright 1998 Red Hat corp --- All Rights Reserved
  *
  * This file is part of the Linux kernel and is made available under
  * the terms of the GNU General Public License, version 2, or at your
  * option, any later version, incorporated herein by reference.
  *
  * Journal commit routines for the generic filesystem journaling code;
  * part of the ext2fs journaling system.
  */
 #include <linux/time.h>
 #include <linux/fs.h>
 #include <linux/jbd.h>
 #include <linux/errno.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
 #include <linux/bio.h>
 /*
  * Default IO end handler for temporary BJ_IO buffer_heads.
  */
 static void journal_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
 {
 	BUFFER_TRACE(bh, "");
 	if (uptodate)
 		set_buffer_uptodate(bh);
 	else
 		clear_buffer_uptodate(bh);
 	unlock_buffer(bh);
 }
 /*
  * When an ext3-ordered file is truncated, it is possible that many pages are
  * not successfully freed, because they are attached to a committing transaction.
  * After the transaction commits, these pages are left on the LRU, with no
  * ->mapping, and with attached buffers.  These pages are trivially reclaimable
  * by the VM, but their apparent absence upsets the VM accounting, and it makes
  * the numbers in /proc/meminfo look odd.
  *
  * So here, we have a buffer which has just come off the forget list.  Look to
  * see if we can strip all buffers from the backing page.
  *
  * Called under journal->j_list_lock.  The caller provided us with a ref
  * against the buffer, and we drop that here.
  */
 static void release_buffer_page(struct buffer_head *bh)
 {
 	struct page *page;
 	if (buffer_dirty(bh))
 		goto nope;
 	if (atomic_read(&bh->b_count) != 1)
 		goto nope;
 	page = bh->b_page;
 	if (!page)
 		goto nope;
 	if (page->mapping)
 		goto nope;
 	/* OK, it's a truncated page */
 	if (!trylock_page(page))
 		goto nope;
 	page_cache_get(page);
 	__brelse(bh);
 	try_to_free_buffers(page);
 	unlock_page(page);
 	page_cache_release(page);
 	return;
 nope:
 	__brelse(bh);
 }
 /*
  * Decrement reference counter for data buffer. If it has been marked
  * 'BH_Freed', release it and the page to which it belongs if possible.
  */
 static void release_data_buffer(struct buffer_head *bh)
 {
 	if (buffer_freed(bh)) {
 		clear_buffer_freed(bh);
 		release_buffer_page(bh);
 	} else
 		put_bh(bh);
 }
 /*
  * Try to acquire jbd_lock_bh_state() against the buffer, when j_list_lock is
  * held.  For ranking reasons we must trylock.  If we lose, schedule away and
  * return 0.  j_list_lock is dropped in this case.
  */
 static int inverted_lock(journal_t *journal, struct buffer_head *bh)
 {
 	if (!jbd_trylock_bh_state(bh)) {
 		spin_unlock(&journal->j_list_lock);
 		schedule();
 		return 0;
 	}
 	return 1;
 }
 /* Done it all: now write the commit record.  We should have
  * cleaned up our previous buffers by now, so if we are in abort
  * mode we can now just skip the rest of the journal write
  * entirely.
  *
  * Returns 1 if the journal needs to be aborted or 0 on success
  */
 static int journal_write_commit_record(journal_t *journal,
 					transaction_t *commit_transaction)
 {
 	struct journal_head *descriptor;
 	struct buffer_head *bh;
 	journal_header_t *header;
 	int ret;
-	int barrier_done = 0;
 	if (is_journal_aborted(journal))
 		return 0;
 	descriptor = journal_get_descriptor_buffer(journal);
 	if (!descriptor)
 		return 1;
 	bh = jh2bh(descriptor);
 	header = (journal_header_t *)(bh->b_data);
 	header->h_magic = cpu_to_be32(JFS_MAGIC_NUMBER);
 	header->h_blocktype = cpu_to_be32(JFS_COMMIT_BLOCK);
 	header->h_sequence = cpu_to_be32(commit_transaction->t_tid);
 	JBUFFER_TRACE(descriptor, "write commit block");
 	set_buffer_dirty(bh);
 	if (journal->j_flags & JFS_BARRIER) {
-		set_buffer_ordered(bh);
+		ret = __sync_dirty_buffer(bh, WRITE_SYNC | WRITE_BARRIER);
-		barrier_done = 1;
-	}
-	ret = sync_dirty_buffer(bh);
-	if (barrier_done)
-		clear_buffer_ordered(bh);
-	/* is it possible for another commit to fail at roughly
-	 * the same time as this one?  If so, we don't want to
-	 * trust the barrier flag in the super, but instead want
-	 * to remember if we sent a barrier request
-	 */
-	if (ret == -EOPNOTSUPP && barrier_done) {
-		char b[BDEVNAME_SIZE];
-		printk(KERN_WARNING
+		/*
-			"JBD: barrier-based sync failed on %s - "
+		 * Is it possible for another commit to fail at roughly
-			"disabling barriers\n",
+		 * the same time as this one?  If so, we don't want to
-			bdevname(journal->j_dev, b));
+		 * trust the barrier flag in the super, but instead want
-		spin_lock(&journal->j_state_lock);
+		 * to remember if we sent a barrier request
-		journal->j_flags &= ~JFS_BARRIER;
+		 */
-		spin_unlock(&journal->j_state_lock);
+		if (ret == -EOPNOTSUPP) {
+			char b[BDEVNAME_SIZE];
-		/* And try again, without the barrier */
+			printk(KERN_WARNING
-		set_buffer_uptodate(bh);
+				"JBD: barrier-based sync failed on %s - "
-		set_buffer_dirty(bh);
+				"disabling barriers\n",
+				bdevname(journal->j_dev, b));
+			spin_lock(&journal->j_state_lock);
+			journal->j_flags &= ~JFS_BARRIER;
+			spin_unlock(&journal->j_state_lock);
+			/* And try again, without the barrier */
+			set_buffer_uptodate(bh);
+			set_buffer_dirty(bh);
+			ret = sync_dirty_buffer(bh);
+		}
+	} else {
 		ret = sync_dirty_buffer(bh);
 	}
 	put_bh(bh);		/* One for getblk() */
 	journal_put_journal_head(descriptor);
 	return (ret == -EIO);
 }
 static void journal_do_submit_data(struct buffer_head **wbuf, int bufs,
 				   int write_op)
 {
 	int i;
 	for (i = 0; i < bufs; i++) {
 		wbuf[i]->b_end_io = end_buffer_write_sync;
 		/* We use-up our safety reference in submit_bh() */
 		submit_bh(write_op, wbuf[i]);
 	}
 }
 /*
  *  Submit all the data buffers to disk
  */
 static int journal_submit_data_buffers(journal_t *journal,
 				       transaction_t *commit_transaction,
 				       int write_op)
 {
 	struct journal_head *jh;
 	struct buffer_head *bh;
 	int locked;
 	int bufs = 0;
 	struct buffer_head **wbuf = journal->j_wbuf;
 	int err = 0;
 	/*
 	 * Whenever we unlock the journal and sleep, things can get added
 	 * onto ->t_sync_datalist, so we have to keep looping back to
 	 * write_out_data until we *know* that the list is empty.
 	 *
 	 * Cleanup any flushed data buffers from the data list.  Even in
 	 * abort mode, we want to flush this out as soon as possible.
 	 */
 write_out_data:
 	cond_resched();
 	spin_lock(&journal->j_list_lock);
 	while (commit_transaction->t_sync_datalist) {
 		jh = commit_transaction->t_sync_datalist;
 		bh = jh2bh(jh);
 		locked = 0;
 		/* Get reference just to make sure buffer does not disappear
 		 * when we are forced to drop various locks */
 		get_bh(bh);
 		/* If the buffer is dirty, we need to submit IO and hence
 		 * we need the buffer lock. We try to lock the buffer without
 		 * blocking. If we fail, we need to drop j_list_lock and do
 		 * blocking lock_buffer().
 		 */
 		if (buffer_dirty(bh)) {
 			if (!trylock_buffer(bh)) {
 				BUFFER_TRACE(bh, "needs blocking lock");
 				spin_unlock(&journal->j_list_lock);
 				/* Write out all data to prevent deadlocks */
 				journal_do_submit_data(wbuf, bufs, write_op);
 				bufs = 0;
 				lock_buffer(bh);
 				spin_lock(&journal->j_list_lock);
 			}
 			locked = 1;
 		}
 		/* We have to get bh_state lock. Again out of order, sigh. */
 		if (!inverted_lock(journal, bh)) {
 			jbd_lock_bh_state(bh);
 			spin_lock(&journal->j_list_lock);
 		}
 		/* Someone already cleaned up the buffer? */
 		if (!buffer_jbd(bh) || bh2jh(bh) != jh
 			|| jh->b_transaction != commit_transaction
 			|| jh->b_jlist != BJ_SyncData) {
 			jbd_unlock_bh_state(bh);
 			if (locked)
 				unlock_buffer(bh);
 			BUFFER_TRACE(bh, "already cleaned up");
 			release_data_buffer(bh);
 			continue;
 		}
 		if (locked && test_clear_buffer_dirty(bh)) {
 			BUFFER_TRACE(bh, "needs writeout, adding to array");
 			wbuf[bufs++] = bh;
 			__journal_file_buffer(jh, commit_transaction,
 						BJ_Locked);
 			jbd_unlock_bh_state(bh);
 			if (bufs == journal->j_wbufsize) {
 				spin_unlock(&journal->j_list_lock);
 				journal_do_submit_data(wbuf, bufs, write_op);
 				bufs = 0;
 				goto write_out_data;
 			}
 		} else if (!locked && buffer_locked(bh)) {
 			__journal_file_buffer(jh, commit_transaction,
 						BJ_Locked);
 			jbd_unlock_bh_state(bh);
 			put_bh(bh);
 		} else {
 			BUFFER_TRACE(bh, "writeout complete: unfile");
 			if (unlikely(!buffer_uptodate(bh)))
 				err = -EIO;
 			__journal_unfile_buffer(jh);
 			jbd_unlock_bh_state(bh);
 			if (locked)
 				unlock_buffer(bh);
 			journal_remove_journal_head(bh);
 			/* One for our safety reference, other for
 			 * journal_remove_journal_head() */
 			put_bh(bh);
 			release_data_buffer(bh);
 		}
 		if (need_resched() || spin_needbreak(&journal->j_list_lock)) {
 			spin_unlock(&journal->j_list_lock);
 			goto write_out_data;
 		}
 	}
 	spin_unlock(&journal->j_list_lock);
 	journal_do_submit_data(wbuf, bufs, write_op);
 	return err;
 }
 /*
  * journal_commit_transaction
  *
  * The primary function for committing a transaction to the log.  This
  * function is called by the journal thread to begin a complete commit.
  */
 void journal_commit_transaction(journal_t *journal)
 {
 	transaction_t *commit_transaction;
 	struct journal_head *jh, *new_jh, *descriptor;
 	struct buffer_head **wbuf = journal->j_wbuf;
 	int bufs;
 	int flags;
 	int err;
 	unsigned int blocknr;
 	ktime_t start_time;
 	u64 commit_time;
 	char *tagp = NULL;
 	journal_header_t *header;
 	journal_block_tag_t *tag = NULL;
 	int space_left = 0;
 	int first_tag = 0;
 	int tag_flag;
 	int i;
 	int write_op = WRITE;
 	/*
 	 * First job: lock down the current transaction and wait for
 	 * all outstanding updates to complete.
 	 */
 #ifdef COMMIT_STATS
 	spin_lock(&journal->j_list_lock);
 	summarise_journal_usage(journal);
 	spin_unlock(&journal->j_list_lock);
 #endif
 	/* Do we need to erase the effects of a prior journal_flush? */
 	if (journal->j_flags & JFS_FLUSHED) {
 		jbd_debug(3, "super block updated\n");
 		journal_update_superblock(journal, 1);
 	} else {
 		jbd_debug(3, "superblock not updated\n");
 	}
 	J_ASSERT(journal->j_running_transaction != NULL);
 	J_ASSERT(journal->j_committing_transaction == NULL);
 	commit_transaction = journal->j_running_transaction;
 	J_ASSERT(commit_transaction->t_state == T_RUNNING);
 	jbd_debug(1, "JBD: starting commit of transaction %d\n",
 			commit_transaction->t_tid);
 	spin_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_LOCKED;
 	/*
 	 * Use plugged writes here, since we want to submit several before
 	 * we unplug the device. We don't do explicit unplugging in here,
 	 * instead we rely on sync_buffer() doing the unplug for us.
 	 */
 	if (commit_transaction->t_synchronous_commit)
 		write_op = WRITE_SYNC_PLUG;
 	spin_lock(&commit_transaction->t_handle_lock);
 	while (commit_transaction->t_updates) {
 		DEFINE_WAIT(wait);
 		prepare_to_wait(&journal->j_wait_updates, &wait,
 					TASK_UNINTERRUPTIBLE);
 		if (commit_transaction->t_updates) {
 			spin_unlock(&commit_transaction->t_handle_lock);
 			spin_unlock(&journal->j_state_lock);
 			schedule();
 			spin_lock(&journal->j_state_lock);
 			spin_lock(&commit_transaction->t_handle_lock);
 		}
 		finish_wait(&journal->j_wait_updates, &wait);
 	}
 	spin_unlock(&commit_transaction->t_handle_lock);
 	J_ASSERT (commit_transaction->t_outstanding_credits <=
 			journal->j_max_transaction_buffers);
 	/*
 	 * First thing we are allowed to do is to discard any remaining
 	 * BJ_Reserved buffers.  Note, it is _not_ permissible to assume
 	 * that there are no such buffers: if a large filesystem
 	 * operation like a truncate needs to split itself over multiple
 	 * transactions, then it may try to do a journal_restart() while
 	 * there are still BJ_Reserved buffers outstanding.  These must
 	 * be released cleanly from the current transaction.
 	 *
 	 * In this case, the filesystem must still reserve write access
 	 * again before modifying the buffer in the new transaction, but
 	 * we do not require it to remember exactly which old buffers it
 	 * has reserved.  This is consistent with the existing behaviour
 	 * that multiple journal_get_write_access() calls to the same
 	 * buffer are perfectly permissable.
 	 */
 	while (commit_transaction->t_reserved_list) {
 		jh = commit_transaction->t_reserved_list;
 		JBUFFER_TRACE(jh, "reserved, unused: refile");
 		/*
 		 * A journal_get_undo_access()+journal_release_buffer() may
 		 * leave undo-committed data.
 		 */
 		if (jh->b_committed_data) {
 			struct buffer_head *bh = jh2bh(jh);
 			jbd_lock_bh_state(bh);
 			jbd_free(jh->b_committed_data, bh->b_size);
 			jh->b_committed_data = NULL;
 			jbd_unlock_bh_state(bh);
 		}
 		journal_refile_buffer(journal, jh);
 	}
 	/*
 	 * Now try to drop any written-back buffers from the journal's
 	 * checkpoint lists.  We do this *before* commit because it potentially
 	 * frees some memory
 	 */
 	spin_lock(&journal->j_list_lock);
 	__journal_clean_checkpoint_list(journal);
 	spin_unlock(&journal->j_list_lock);
 	jbd_debug (3, "JBD: commit phase 1\n");
 	/*
 	 * Switch to a new revoke table.
 	 */
 	journal_switch_revoke_table(journal);
 	commit_transaction->t_state = T_FLUSH;
 	journal->j_committing_transaction = commit_transaction;
 	journal->j_running_transaction = NULL;
 	start_time = ktime_get();
 	commit_transaction->t_log_start = journal->j_head;
 	wake_up(&journal->j_wait_transaction_locked);
 	spin_unlock(&journal->j_state_lock);
 	jbd_debug (3, "JBD: commit phase 2\n");
 	/*
 	 * Now start flushing things to disk, in the order they appear
 	 * on the transaction lists.  Data blocks go first.
 	 */
 	err = journal_submit_data_buffers(journal, commit_transaction,
 					  write_op);
 	/*
 	 * Wait for all previously submitted IO to complete.
 	 */
 	spin_lock(&journal->j_list_lock);
 	while (commit_transaction->t_locked_list) {
 		struct buffer_head *bh;
 		jh = commit_transaction->t_locked_list->b_tprev;
 		bh = jh2bh(jh);
 		get_bh(bh);
 		if (buffer_locked(bh)) {
 			spin_unlock(&journal->j_list_lock);
 			wait_on_buffer(bh);
 			spin_lock(&journal->j_list_lock);
 		}
 		if (unlikely(!buffer_uptodate(bh))) {
 			if (!trylock_page(bh->b_page)) {
 				spin_unlock(&journal->j_list_lock);
 				lock_page(bh->b_page);
 				spin_lock(&journal->j_list_lock);
 			}
 			if (bh->b_page->mapping)
 				set_bit(AS_EIO, &bh->b_page->mapping->flags);
 			unlock_page(bh->b_page);
 			SetPageError(bh->b_page);
 			err = -EIO;
 		}
 		if (!inverted_lock(journal, bh)) {
 			put_bh(bh);
 			spin_lock(&journal->j_list_lock);
 			continue;
 		}
 		if (buffer_jbd(bh) && bh2jh(bh) == jh &&
 		    jh->b_transaction == commit_transaction &&
 		    jh->b_jlist == BJ_Locked) {
 			__journal_unfile_buffer(jh);
 			jbd_unlock_bh_state(bh);
 			journal_remove_journal_head(bh);
 			put_bh(bh);
 		} else {
 			jbd_unlock_bh_state(bh);
 		}
 		release_data_buffer(bh);
 		cond_resched_lock(&journal->j_list_lock);
 	}
 	spin_unlock(&journal->j_list_lock);
 	if (err) {
 		char b[BDEVNAME_SIZE];
 		printk(KERN_WARNING
 			"JBD: Detected IO errors while flushing file data "
 			"on %s\n", bdevname(journal->j_fs_dev, b));
 		if (journal->j_flags & JFS_ABORT_ON_SYNCDATA_ERR)
 			journal_abort(journal, err);
 		err = 0;
 	}
 	journal_write_revoke_records(journal, commit_transaction, write_op);
 	/*
 	 * If we found any dirty or locked buffers, then we should have
 	 * looped back up to the write_out_data label.  If there weren't
 	 * any then journal_clean_data_list should have wiped the list
 	 * clean by now, so check that it is in fact empty.
 	 */
 	J_ASSERT (commit_transaction->t_sync_datalist == NULL);
 	jbd_debug (3, "JBD: commit phase 3\n");
 	/*
 	 * Way to go: we have now written out all of the data for a
 	 * transaction!  Now comes the tricky part: we need to write out
 	 * metadata.  Loop over the transaction's entire buffer list:
 	 */
 	spin_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_COMMIT;
 	spin_unlock(&journal->j_state_lock);
 	J_ASSERT(commit_transaction->t_nr_buffers <=
 		 commit_transaction->t_outstanding_credits);
 	descriptor = NULL;
 	bufs = 0;
 	while (commit_transaction->t_buffers) {
 		/* Find the next buffer to be journaled... */
 		jh = commit_transaction->t_buffers;
 		/* If we're in abort mode, we just un-journal the buffer and
 		   release it. */
 		if (is_journal_aborted(journal)) {
 			clear_buffer_jbddirty(jh2bh(jh));
 			JBUFFER_TRACE(jh, "journal is aborting: refile");
 			journal_refile_buffer(journal, jh);
 			/* If that was the last one, we need to clean up
 			 * any descriptor buffers which may have been
 			 * already allocated, even if we are now
 			 * aborting. */
 			if (!commit_transaction->t_buffers)
 				goto start_journal_io;
 			continue;
 		}
 		/* Make sure we have a descriptor block in which to
 		   record the metadata buffer. */
 		if (!descriptor) {
 			struct buffer_head *bh;
 			J_ASSERT (bufs == 0);
 			jbd_debug(4, "JBD: get descriptor\n");
 			descriptor = journal_get_descriptor_buffer(journal);
 			if (!descriptor) {
 				journal_abort(journal, -EIO);
 				continue;
 			}
 			bh = jh2bh(descriptor);
 			jbd_debug(4, "JBD: got buffer %llu (%p)\n",
 				(unsigned long long)bh->b_blocknr, bh->b_data);
 			header = (journal_header_t *)&bh->b_data[0];
 			header->h_magic     = cpu_to_be32(JFS_MAGIC_NUMBER);
 			header->h_blocktype = cpu_to_be32(JFS_DESCRIPTOR_BLOCK);
 			header->h_sequence  = cpu_to_be32(commit_transaction->t_tid);
 			tagp = &bh->b_data[sizeof(journal_header_t)];
 			space_left = bh->b_size - sizeof(journal_header_t);
 			first_tag = 1;
 			set_buffer_jwrite(bh);
 			set_buffer_dirty(bh);
 			wbuf[bufs++] = bh;
 			/* Record it so that we can wait for IO
                            completion later */
 			BUFFER_TRACE(bh, "ph3: file as descriptor");
 			journal_file_buffer(descriptor, commit_transaction,
 					BJ_LogCtl);
 		}
 		/* Where is the buffer to be written? */
 		err = journal_next_log_block(journal, &blocknr);
 		/* If the block mapping failed, just abandon the buffer
 		   and repeat this loop: we'll fall into the
 		   refile-on-abort condition above. */
 		if (err) {
 			journal_abort(journal, err);
 			continue;
 		}
 		/*
 		 * start_this_handle() uses t_outstanding_credits to determine
 		 * the free space in the log, but this counter is changed
 		 * by journal_next_log_block() also.
 		 */
 		commit_transaction->t_outstanding_credits--;
 		/* Bump b_count to prevent truncate from stumbling over
                    the shadowed buffer!  @@@ This can go if we ever get
                    rid of the BJ_IO/BJ_Shadow pairing of buffers. */
 		atomic_inc(&jh2bh(jh)->b_count);
 		/* Make a temporary IO buffer with which to write it out
                    (this will requeue both the metadata buffer and the
                    temporary IO buffer). new_bh goes on BJ_IO*/
 		set_bit(BH_JWrite, &jh2bh(jh)->b_state);
 		/*
 		 * akpm: journal_write_metadata_buffer() sets
 		 * new_bh->b_transaction to commit_transaction.
 		 * We need to clean this up before we release new_bh
 		 * (which is of type BJ_IO)
 		 */
 		JBUFFER_TRACE(jh, "ph3: write metadata");
 		flags = journal_write_metadata_buffer(commit_transaction,
 						      jh, &new_jh, blocknr);
 		set_bit(BH_JWrite, &jh2bh(new_jh)->b_state);
 		wbuf[bufs++] = jh2bh(new_jh);
 		/* Record the new block's tag in the current descriptor
                    buffer */
 		tag_flag = 0;
 		if (flags & 1)
 			tag_flag |= JFS_FLAG_ESCAPE;
 		if (!first_tag)
 			tag_flag |= JFS_FLAG_SAME_UUID;
 		tag = (journal_block_tag_t *) tagp;
 		tag->t_blocknr = cpu_to_be32(jh2bh(jh)->b_blocknr);
 		tag->t_flags = cpu_to_be32(tag_flag);
 		tagp += sizeof(journal_block_tag_t);
 		space_left -= sizeof(journal_block_tag_t);
 		if (first_tag) {
 			memcpy (tagp, journal->j_uuid, 16);
 			tagp += 16;
 			space_left -= 16;
 			first_tag = 0;
 		}
 		/* If there's no more to do, or if the descriptor is full,
 		   let the IO rip! */
 		if (bufs == journal->j_wbufsize ||
 		    commit_transaction->t_buffers == NULL ||
 		    space_left < sizeof(journal_block_tag_t) + 16) {
 			jbd_debug(4, "JBD: Submit %d IOs\n", bufs);
 			/* Write an end-of-descriptor marker before
                            submitting the IOs.  "tag" still points to
                            the last tag we set up. */
 			tag->t_flags |= cpu_to_be32(JFS_FLAG_LAST_TAG);
 start_journal_io:
 			for (i = 0; i < bufs; i++) {
 				struct buffer_head *bh = wbuf[i];
 				lock_buffer(bh);
 				clear_buffer_dirty(bh);
 				set_buffer_uptodate(bh);
 				bh->b_end_io = journal_end_buffer_io_sync;
 				submit_bh(write_op, bh);
 			}
 			cond_resched();
 			/* Force a new descriptor to be generated next
                            time round the loop. */
 			descriptor = NULL;
 			bufs = 0;
 		}
 	}
 	/* Lo and behold: we have just managed to send a transaction to
            the log.  Before we can commit it, wait for the IO so far to
            complete.  Control buffers being written are on the
            transaction's t_log_list queue, and metadata buffers are on
            the t_iobuf_list queue.
 	   Wait for the buffers in reverse order.  That way we are
 	   less likely to be woken up until all IOs have completed, and
 	   so we incur less scheduling load.
 	*/
 	jbd_debug(3, "JBD: commit phase 4\n");
 	/*
 	 * akpm: these are BJ_IO, and j_list_lock is not needed.
 	 * See __journal_try_to_free_buffer.
 	 */
 wait_for_iobuf:
 	while (commit_transaction->t_iobuf_list != NULL) {
 		struct buffer_head *bh;
 		jh = commit_transaction->t_iobuf_list->b_tprev;
 		bh = jh2bh(jh);
 		if (buffer_locked(bh)) {
 			wait_on_buffer(bh);
 			goto wait_for_iobuf;
 		}
 		if (cond_resched())
 			goto wait_for_iobuf;
 		if (unlikely(!buffer_uptodate(bh)))
 			err = -EIO;
 		clear_buffer_jwrite(bh);
 		JBUFFER_TRACE(jh, "ph4: unfile after journal write");
 		journal_unfile_buffer(journal, jh);
 		/*
 		 * ->t_iobuf_list should contain only dummy buffer_heads
 		 * which were created by journal_write_metadata_buffer().
 		 */
 		BUFFER_TRACE(bh, "dumping temporary bh");
 		journal_put_journal_head(jh);
 		__brelse(bh);
 		J_ASSERT_BH(bh, atomic_read(&bh->b_count) == 0);
 		free_buffer_head(bh);
 		/* We also have to unlock and free the corresponding
                    shadowed buffer */
 		jh = commit_transaction->t_shadow_list->b_tprev;
 		bh = jh2bh(jh);
 		clear_bit(BH_JWrite, &bh->b_state);
 		J_ASSERT_BH(bh, buffer_jbddirty(bh));
 		/* The metadata is now released for reuse, but we need
                    to remember it against this transaction so that when
                    we finally commit, we can do any checkpointing
                    required. */
 		JBUFFER_TRACE(jh, "file as BJ_Forget");
 		journal_file_buffer(jh, commit_transaction, BJ_Forget);
 		/* Wake up any transactions which were waiting for this
 		   IO to complete */
 		wake_up_bit(&bh->b_state, BH_Unshadow);
 		JBUFFER_TRACE(jh, "brelse shadowed buffer");
 		__brelse(bh);
 	}
 	J_ASSERT (commit_transaction->t_shadow_list == NULL);
 	jbd_debug(3, "JBD: commit phase 5\n");
 	/* Here we wait for the revoke record and descriptor record buffers */
  wait_for_ctlbuf:
 	while (commit_transaction->t_log_list != NULL) {
 		struct buffer_head *bh;
 		jh = commit_transaction->t_log_list->b_tprev;
 		bh = jh2bh(jh);
 		if (buffer_locked(bh)) {
 			wait_on_buffer(bh);
 			goto wait_for_ctlbuf;
 		}
 		if (cond_resched())
 			goto wait_for_ctlbuf;
 		if (unlikely(!buffer_uptodate(bh)))
 			err = -EIO;
 		BUFFER_TRACE(bh, "ph5: control buffer writeout done: unfile");
 		clear_buffer_jwrite(bh);
 		journal_unfile_buffer(journal, jh);
 		journal_put_journal_head(jh);
 		__brelse(bh);		/* One for getblk */
 		/* AKPM: bforget here */
 	}
 	if (err)
 		journal_abort(journal, err);
 	jbd_debug(3, "JBD: commit phase 6\n");
 	/* All metadata is written, now write commit record and do cleanup */
 	spin_lock(&journal->j_state_lock);
 	J_ASSERT(commit_transaction->t_state == T_COMMIT);
 	commit_transaction->t_state = T_COMMIT_RECORD;
 	spin_unlock(&journal->j_state_lock);
 	if (journal_write_commit_record(journal, commit_transaction))
 		err = -EIO;
 	if (err)
 		journal_abort(journal, err);
 	/* End of a transaction!  Finally, we can do checkpoint
            processing: any buffers committed as a result of this
            transaction can be removed from any checkpoint list it was on
            before. */
 	jbd_debug(3, "JBD: commit phase 7\n");
 	J_ASSERT(commit_transaction->t_sync_datalist == NULL);
 	J_ASSERT(commit_transaction->t_buffers == NULL);
 	J_ASSERT(commit_transaction->t_checkpoint_list == NULL);
 	J_ASSERT(commit_transaction->t_iobuf_list == NULL);
 	J_ASSERT(commit_transaction->t_shadow_list == NULL);
 	J_ASSERT(commit_transaction->t_log_list == NULL);
 restart_loop:
 	/*
 	 * As there are other places (journal_unmap_buffer()) adding buffers
 	 * to this list we have to be careful and hold the j_list_lock.
 	 */
 	spin_lock(&journal->j_list_lock);
 	while (commit_transaction->t_forget) {
 		transaction_t *cp_transaction;
 		struct buffer_head *bh;
 		jh = commit_transaction->t_forget;
 		spin_unlock(&journal->j_list_lock);
 		bh = jh2bh(jh);
 		jbd_lock_bh_state(bh);
 		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction ||
 			jh->b_transaction == journal->j_running_transaction);
 		/*
 		 * If there is undo-protected committed data against
 		 * this buffer, then we can remove it now.  If it is a
 		 * buffer needing such protection, the old frozen_data
 		 * field now points to a committed version of the
 		 * buffer, so rotate that field to the new committed
 		 * data.
 		 *
 		 * Otherwise, we can just throw away the frozen data now.
 		 */
 		if (jh->b_committed_data) {
 			jbd_free(jh->b_committed_data, bh->b_size);
 			jh->b_committed_data = NULL;
 			if (jh->b_frozen_data) {
 				jh->b_committed_data = jh->b_frozen_data;
 				jh->b_frozen_data = NULL;
 			}
 		} else if (jh->b_frozen_data) {
 			jbd_free(jh->b_frozen_data, bh->b_size);
 			jh->b_frozen_data = NULL;
 		}
 		spin_lock(&journal->j_list_lock);
 		cp_transaction = jh->b_cp_transaction;
 		if (cp_transaction) {
 			JBUFFER_TRACE(jh, "remove from old cp transaction");
 			__journal_remove_checkpoint(jh);
 		}
 		/* Only re-checkpoint the buffer_head if it is marked
 		 * dirty.  If the buffer was added to the BJ_Forget list
 		 * by journal_forget, it may no longer be dirty and
 		 * there's no point in keeping a checkpoint record for
 		 * it. */
 		/* A buffer which has been freed while still being
 		 * journaled by a previous transaction may end up still
 		 * being dirty here, but we want to avoid writing back
 		 * that buffer in the future after the "add to orphan"
 		 * operation been committed,  That's not only a performance
 		 * gain, it also stops aliasing problems if the buffer is
 		 * left behind for writeback and gets reallocated for another
 		 * use in a different page. */
 		if (buffer_freed(bh) && !jh->b_next_transaction) {
 			clear_buffer_freed(bh);
 			clear_buffer_jbddirty(bh);
 		}
 		if (buffer_jbddirty(bh)) {
 			JBUFFER_TRACE(jh, "add to new checkpointing trans");
 			__journal_insert_checkpoint(jh, commit_transaction);
 			if (is_journal_aborted(journal))
 				clear_buffer_jbddirty(bh);
 			JBUFFER_TRACE(jh, "refile for checkpoint writeback");
 			__journal_refile_buffer(jh);
 			jbd_unlock_bh_state(bh);
 		} else {
 			J_ASSERT_BH(bh, !buffer_dirty(bh));
 			/* The buffer on BJ_Forget list and not jbddirty means
 			 * it has been freed by this transaction and hence it
 			 * could not have been reallocated until this
 			 * transaction has committed. *BUT* it could be
 			 * reallocated once we have written all the data to
 			 * disk and before we process the buffer on BJ_Forget
 			 * list. */
 			JBUFFER_TRACE(jh, "refile or unfile freed buffer");
 			__journal_refile_buffer(jh);
 			if (!jh->b_transaction) {
 				jbd_unlock_bh_state(bh);
 				 /* needs a brelse */
 				journal_remove_journal_head(bh);
 				release_buffer_page(bh);
 			} else
 				jbd_unlock_bh_state(bh);
 		}
 		cond_resched_lock(&journal->j_list_lock);
 	}
 	spin_unlock(&journal->j_list_lock);
 	/*
 	 * This is a bit sleazy.  We use j_list_lock to protect transition
 	 * of a transaction into T_FINISHED state and calling
 	 * __journal_drop_transaction(). Otherwise we could race with
 	 * other checkpointing code processing the transaction...
 	 */
 	spin_lock(&journal->j_state_lock);
 	spin_lock(&journal->j_list_lock);
 	/*
 	 * Now recheck if some buffers did not get attached to the transaction
 	 * while the lock was dropped...
 	 */
 	if (commit_transaction->t_forget) {
 		spin_unlock(&journal->j_list_lock);
 		spin_unlock(&journal->j_state_lock);
 		goto restart_loop;
 	}
 	/* Done with this transaction! */
 	jbd_debug(3, "JBD: commit phase 8\n");
 	J_ASSERT(commit_transaction->t_state == T_COMMIT_RECORD);
 	commit_transaction->t_state = T_FINISHED;
 	J_ASSERT(commit_transaction == journal->j_committing_transaction);
 	journal->j_commit_sequence = commit_transaction->t_tid;
 	journal->j_committing_transaction = NULL;
 	commit_time = ktime_to_ns(ktime_sub(ktime_get(), start_time));
 	/*
 	 * weight the commit time higher than the average time so we don't
 	 * react too strongly to vast changes in commit time
 	 */
 	if (likely(journal->j_average_commit_time))
 		journal->j_average_commit_time = (commit_time*3 +
 				journal->j_average_commit_time) / 4;
 	else
 		journal->j_average_commit_time = commit_time;
 	spin_unlock(&journal->j_state_lock);
 	if (commit_transaction->t_checkpoint_list == NULL &&
 	    commit_transaction->t_checkpoint_io_list == NULL) {
 		__journal_drop_transaction(journal, commit_transaction);
 	} else {
 		if (journal->j_checkpoint_transactions == NULL) {
 			journal->j_checkpoint_transactions = commit_transaction;
 			commit_transaction->t_cpnext = commit_transaction;
 			commit_transaction->t_cpprev = commit_transaction;
 		} else {
 			commit_transaction->t_cpnext =
 				journal->j_checkpoint_transactions;
 			commit_transaction->t_cpprev =
 				commit_transaction->t_cpnext->t_cpprev;
 			commit_transaction->t_cpnext->t_cpprev =

fs/jbd2/commit.c

Diff comments View file @ 87e9951

 /*
  * linux/fs/jbd2/commit.c
  *
  * Written by Stephen C. Tweedie <sct@redhat.com>, 1998
  *
  * Copyright 1998 Red Hat corp --- All Rights Reserved
  *
  * This file is part of the Linux kernel and is made available under
  * the terms of the GNU General Public License, version 2, or at your
  * option, any later version, incorporated herein by reference.
  *
  * Journal commit routines for the generic filesystem journaling code;
  * part of the ext2fs journaling system.
  */
 #include <linux/time.h>
 #include <linux/fs.h>
 #include <linux/jbd2.h>
 #include <linux/errno.h>
 #include <linux/slab.h>
 #include <linux/mm.h>
 #include <linux/pagemap.h>
 #include <linux/jiffies.h>
 #include <linux/crc32.h>
 #include <linux/writeback.h>
 #include <linux/backing-dev.h>
 #include <linux/bio.h>
 #include <linux/blkdev.h>
 #include <trace/events/jbd2.h>
 /*
  * Default IO end handler for temporary BJ_IO buffer_heads.
  */
 static void journal_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
 {
 	BUFFER_TRACE(bh, "");
 	if (uptodate)
 		set_buffer_uptodate(bh);
 	else
 		clear_buffer_uptodate(bh);
 	unlock_buffer(bh);
 }
 /*
  * When an ext4 file is truncated, it is possible that some pages are not
  * successfully freed, because they are attached to a committing transaction.
  * After the transaction commits, these pages are left on the LRU, with no
  * ->mapping, and with attached buffers.  These pages are trivially reclaimable
  * by the VM, but their apparent absence upsets the VM accounting, and it makes
  * the numbers in /proc/meminfo look odd.
  *
  * So here, we have a buffer which has just come off the forget list.  Look to
  * see if we can strip all buffers from the backing page.
  *
  * Called under lock_journal(), and possibly under journal_datalist_lock.  The
  * caller provided us with a ref against the buffer, and we drop that here.
  */
 static void release_buffer_page(struct buffer_head *bh)
 {
 	struct page *page;
 	if (buffer_dirty(bh))
 		goto nope;
 	if (atomic_read(&bh->b_count) != 1)
 		goto nope;
 	page = bh->b_page;
 	if (!page)
 		goto nope;
 	if (page->mapping)
 		goto nope;
 	/* OK, it's a truncated page */
 	if (!trylock_page(page))
 		goto nope;
 	page_cache_get(page);
 	__brelse(bh);
 	try_to_free_buffers(page);
 	unlock_page(page);
 	page_cache_release(page);
 	return;
 nope:
 	__brelse(bh);
 }
 /*
  * Done it all: now submit the commit record.  We should have
  * cleaned up our previous buffers by now, so if we are in abort
  * mode we can now just skip the rest of the journal write
  * entirely.
  *
  * Returns 1 if the journal needs to be aborted or 0 on success
  */
 static int journal_submit_commit_record(journal_t *journal,
 					transaction_t *commit_transaction,
 					struct buffer_head **cbh,
 					__u32 crc32_sum)
 {
 	struct journal_head *descriptor;
 	struct commit_header *tmp;
 	struct buffer_head *bh;
 	int ret;
-	int barrier_done = 0;
 	struct timespec now = current_kernel_time();
 	if (is_journal_aborted(journal))
 		return 0;
 	descriptor = jbd2_journal_get_descriptor_buffer(journal);
 	if (!descriptor)
 		return 1;
 	bh = jh2bh(descriptor);
 	tmp = (struct commit_header *)bh->b_data;
 	tmp->h_magic = cpu_to_be32(JBD2_MAGIC_NUMBER);
 	tmp->h_blocktype = cpu_to_be32(JBD2_COMMIT_BLOCK);
 	tmp->h_sequence = cpu_to_be32(commit_transaction->t_tid);
 	tmp->h_commit_sec = cpu_to_be64(now.tv_sec);
 	tmp->h_commit_nsec = cpu_to_be32(now.tv_nsec);
 	if (JBD2_HAS_COMPAT_FEATURE(journal,
 				    JBD2_FEATURE_COMPAT_CHECKSUM)) {
 		tmp->h_chksum_type 	= JBD2_CRC32_CHKSUM;
 		tmp->h_chksum_size 	= JBD2_CRC32_CHKSUM_SIZE;
 		tmp->h_chksum[0] 	= cpu_to_be32(crc32_sum);
 	}
 	JBUFFER_TRACE(descriptor, "submit commit block");
 	lock_buffer(bh);
 	clear_buffer_dirty(bh);
 	set_buffer_uptodate(bh);
 	bh->b_end_io = journal_end_buffer_io_sync;
 	if (journal->j_flags & JBD2_BARRIER &&
 	    !JBD2_HAS_INCOMPAT_FEATURE(journal,
 				       JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) {
-		set_buffer_ordered(bh);
+		ret = submit_bh(WRITE_SYNC_PLUG | WRITE_BARRIER, bh);
-		barrier_done = 1;
+		if (ret == -EOPNOTSUPP) {
-	}
+			printk(KERN_WARNING
-	ret = submit_bh(WRITE_SYNC_PLUG, bh);
+			       "JBD2: Disabling barriers on %s, "
-	if (barrier_done)
+			       "not supported by device\n", journal->j_devname);
-		clear_buffer_ordered(bh);
+			write_lock(&journal->j_state_lock);
+			journal->j_flags &= ~JBD2_BARRIER;
+			write_unlock(&journal->j_state_lock);
-	/* is it possible for another commit to fail at roughly
+			/* And try again, without the barrier */
-	 * the same time as this one?  If so, we don't want to
+			lock_buffer(bh);
-	 * trust the barrier flag in the super, but instead want
+			set_buffer_uptodate(bh);
-	 * to remember if we sent a barrier request
+			clear_buffer_dirty(bh);
-	 */
+			ret = submit_bh(WRITE_SYNC_PLUG, bh);
-	if (ret == -EOPNOTSUPP && barrier_done) {
+		}
-		printk(KERN_WARNING
+	} else {
-		       "JBD2: Disabling barriers on %s, "
-		       "not supported by device\n", journal->j_devname);
-		write_lock(&journal->j_state_lock);
-		journal->j_flags &= ~JBD2_BARRIER;
-		write_unlock(&journal->j_state_lock);
-		/* And try again, without the barrier */
-		lock_buffer(bh);
-		set_buffer_uptodate(bh);
-		clear_buffer_dirty(bh);
 		ret = submit_bh(WRITE_SYNC_PLUG, bh);
 	}
 	*cbh = bh;
 	return ret;
 }
 /*
  * This function along with journal_submit_commit_record
  * allows to write the commit record asynchronously.
  */
 static int journal_wait_on_commit_record(journal_t *journal,
 					 struct buffer_head *bh)
 {
 	int ret = 0;
 retry:
 	clear_buffer_dirty(bh);
 	wait_on_buffer(bh);
 	if (buffer_eopnotsupp(bh) && (journal->j_flags & JBD2_BARRIER)) {
 		printk(KERN_WARNING
 		       "JBD2: %s: disabling barries on %s - not supported "
 		       "by device\n", __func__, journal->j_devname);
 		write_lock(&journal->j_state_lock);
 		journal->j_flags &= ~JBD2_BARRIER;
 		write_unlock(&journal->j_state_lock);
 		lock_buffer(bh);
 		clear_buffer_dirty(bh);
 		set_buffer_uptodate(bh);
 		bh->b_end_io = journal_end_buffer_io_sync;
 		ret = submit_bh(WRITE_SYNC_PLUG, bh);
 		if (ret) {
 			unlock_buffer(bh);
 			return ret;
 		}
 		goto retry;
 	}
 	if (unlikely(!buffer_uptodate(bh)))
 		ret = -EIO;
 	put_bh(bh);            /* One for getblk() */
 	jbd2_journal_put_journal_head(bh2jh(bh));
 	return ret;
 }
 /*
  * write the filemap data using writepage() address_space_operations.
  * We don't do block allocation here even for delalloc. We don't
  * use writepages() because with dealyed allocation we may be doing
  * block allocation in writepages().
  */
 static int journal_submit_inode_data_buffers(struct address_space *mapping)
 {
 	int ret;
 	struct writeback_control wbc = {
 		.sync_mode =  WB_SYNC_ALL,
 		.nr_to_write = mapping->nrpages * 2,
 		.range_start = 0,
 		.range_end = i_size_read(mapping->host),
 	};
 	ret = generic_writepages(mapping, &wbc);
 	return ret;
 }
 /*
  * Submit all the data buffers of inode associated with the transaction to
  * disk.
  *
  * We are in a committing transaction. Therefore no new inode can be added to
  * our inode list. We use JI_COMMIT_RUNNING flag to protect inode we currently
  * operate on from being released while we write out pages.
  */
 static int journal_submit_data_buffers(journal_t *journal,
 		transaction_t *commit_transaction)
 {
 	struct jbd2_inode *jinode;
 	int err, ret = 0;
 	struct address_space *mapping;
 	spin_lock(&journal->j_list_lock);
 	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
 		mapping = jinode->i_vfs_inode->i_mapping;
 		jinode->i_flags |= JI_COMMIT_RUNNING;
 		spin_unlock(&journal->j_list_lock);
 		/*
 		 * submit the inode data buffers. We use writepage
 		 * instead of writepages. Because writepages can do
 		 * block allocation  with delalloc. We need to write
 		 * only allocated blocks here.
 		 */
 		trace_jbd2_submit_inode_data(jinode->i_vfs_inode);
 		err = journal_submit_inode_data_buffers(mapping);
 		if (!ret)
 			ret = err;
 		spin_lock(&journal->j_list_lock);
 		J_ASSERT(jinode->i_transaction == commit_transaction);
 		commit_transaction->t_flushed_data_blocks = 1;
 		jinode->i_flags &= ~JI_COMMIT_RUNNING;
 		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
 	}
 	spin_unlock(&journal->j_list_lock);
 	return ret;
 }
 /*
  * Wait for data submitted for writeout, refile inodes to proper
  * transaction if needed.
  *
  */
 static int journal_finish_inode_data_buffers(journal_t *journal,
 		transaction_t *commit_transaction)
 {
 	struct jbd2_inode *jinode, *next_i;
 	int err, ret = 0;
 	/* For locking, see the comment in journal_submit_data_buffers() */
 	spin_lock(&journal->j_list_lock);
 	list_for_each_entry(jinode, &commit_transaction->t_inode_list, i_list) {
 		jinode->i_flags |= JI_COMMIT_RUNNING;
 		spin_unlock(&journal->j_list_lock);
 		err = filemap_fdatawait(jinode->i_vfs_inode->i_mapping);
 		if (err) {
 			/*
 			 * Because AS_EIO is cleared by
 			 * filemap_fdatawait_range(), set it again so
 			 * that user process can get -EIO from fsync().
 			 */
 			set_bit(AS_EIO,
 				&jinode->i_vfs_inode->i_mapping->flags);
 			if (!ret)
 				ret = err;
 		}
 		spin_lock(&journal->j_list_lock);
 		jinode->i_flags &= ~JI_COMMIT_RUNNING;
 		wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING);
 	}
 	/* Now refile inode to proper lists */
 	list_for_each_entry_safe(jinode, next_i,
 				 &commit_transaction->t_inode_list, i_list) {
 		list_del(&jinode->i_list);
 		if (jinode->i_next_transaction) {
 			jinode->i_transaction = jinode->i_next_transaction;
 			jinode->i_next_transaction = NULL;
 			list_add(&jinode->i_list,
 				&jinode->i_transaction->t_inode_list);
 		} else {
 			jinode->i_transaction = NULL;
 		}
 	}
 	spin_unlock(&journal->j_list_lock);
 	return ret;
 }
 static __u32 jbd2_checksum_data(__u32 crc32_sum, struct buffer_head *bh)
 {
 	struct page *page = bh->b_page;
 	char *addr;
 	__u32 checksum;
 	addr = kmap_atomic(page, KM_USER0);
 	checksum = crc32_be(crc32_sum,
 		(void *)(addr + offset_in_page(bh->b_data)), bh->b_size);
 	kunmap_atomic(addr, KM_USER0);
 	return checksum;
 }
 static void write_tag_block(int tag_bytes, journal_block_tag_t *tag,
 				   unsigned long long block)
 {
 	tag->t_blocknr = cpu_to_be32(block & (u32)~0);
 	if (tag_bytes > JBD2_TAG_SIZE32)
 		tag->t_blocknr_high = cpu_to_be32((block >> 31) >> 1);
 }
 /*
  * jbd2_journal_commit_transaction
  *
  * The primary function for committing a transaction to the log.  This
  * function is called by the journal thread to begin a complete commit.
  */
 void jbd2_journal_commit_transaction(journal_t *journal)
 {
 	struct transaction_stats_s stats;
 	transaction_t *commit_transaction;
 	struct journal_head *jh, *new_jh, *descriptor;
 	struct buffer_head **wbuf = journal->j_wbuf;
 	int bufs;
 	int flags;
 	int err;
 	unsigned long long blocknr;
 	ktime_t start_time;
 	u64 commit_time;
 	char *tagp = NULL;
 	journal_header_t *header;
 	journal_block_tag_t *tag = NULL;
 	int space_left = 0;
 	int first_tag = 0;
 	int tag_flag;
 	int i, to_free = 0;
 	int tag_bytes = journal_tag_bytes(journal);
 	struct buffer_head *cbh = NULL; /* For transactional checksums */
 	__u32 crc32_sum = ~0;
 	int write_op = WRITE;
 	/*
 	 * First job: lock down the current transaction and wait for
 	 * all outstanding updates to complete.
 	 */
 #ifdef COMMIT_STATS
 	spin_lock(&journal->j_list_lock);
 	summarise_journal_usage(journal);
 	spin_unlock(&journal->j_list_lock);
 #endif
 	/* Do we need to erase the effects of a prior jbd2_journal_flush? */
 	if (journal->j_flags & JBD2_FLUSHED) {
 		jbd_debug(3, "super block updated\n");
 		jbd2_journal_update_superblock(journal, 1);
 	} else {
 		jbd_debug(3, "superblock not updated\n");
 	}
 	J_ASSERT(journal->j_running_transaction != NULL);
 	J_ASSERT(journal->j_committing_transaction == NULL);
 	commit_transaction = journal->j_running_transaction;
 	J_ASSERT(commit_transaction->t_state == T_RUNNING);
 	trace_jbd2_start_commit(journal, commit_transaction);
 	jbd_debug(1, "JBD: starting commit of transaction %d\n",
 			commit_transaction->t_tid);
 	write_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_LOCKED;
 	/*
 	 * Use plugged writes here, since we want to submit several before
 	 * we unplug the device. We don't do explicit unplugging in here,
 	 * instead we rely on sync_buffer() doing the unplug for us.
 	 */
 	if (commit_transaction->t_synchronous_commit)
 		write_op = WRITE_SYNC_PLUG;
 	trace_jbd2_commit_locking(journal, commit_transaction);
 	stats.run.rs_wait = commit_transaction->t_max_wait;
 	stats.run.rs_locked = jiffies;
 	stats.run.rs_running = jbd2_time_diff(commit_transaction->t_start,
 					      stats.run.rs_locked);
 	spin_lock(&commit_transaction->t_handle_lock);
 	while (atomic_read(&commit_transaction->t_updates)) {
 		DEFINE_WAIT(wait);
 		prepare_to_wait(&journal->j_wait_updates, &wait,
 					TASK_UNINTERRUPTIBLE);
 		if (atomic_read(&commit_transaction->t_updates)) {
 			spin_unlock(&commit_transaction->t_handle_lock);
 			write_unlock(&journal->j_state_lock);
 			schedule();
 			write_lock(&journal->j_state_lock);
 			spin_lock(&commit_transaction->t_handle_lock);
 		}
 		finish_wait(&journal->j_wait_updates, &wait);
 	}
 	spin_unlock(&commit_transaction->t_handle_lock);
 	J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
 			journal->j_max_transaction_buffers);
 	/*
 	 * First thing we are allowed to do is to discard any remaining
 	 * BJ_Reserved buffers.  Note, it is _not_ permissible to assume
 	 * that there are no such buffers: if a large filesystem
 	 * operation like a truncate needs to split itself over multiple
 	 * transactions, then it may try to do a jbd2_journal_restart() while
 	 * there are still BJ_Reserved buffers outstanding.  These must
 	 * be released cleanly from the current transaction.
 	 *
 	 * In this case, the filesystem must still reserve write access
 	 * again before modifying the buffer in the new transaction, but
 	 * we do not require it to remember exactly which old buffers it
 	 * has reserved.  This is consistent with the existing behaviour
 	 * that multiple jbd2_journal_get_write_access() calls to the same
 	 * buffer are perfectly permissable.
 	 */
 	while (commit_transaction->t_reserved_list) {
 		jh = commit_transaction->t_reserved_list;
 		JBUFFER_TRACE(jh, "reserved, unused: refile");
 		/*
 		 * A jbd2_journal_get_undo_access()+jbd2_journal_release_buffer() may
 		 * leave undo-committed data.
 		 */
 		if (jh->b_committed_data) {
 			struct buffer_head *bh = jh2bh(jh);
 			jbd_lock_bh_state(bh);
 			jbd2_free(jh->b_committed_data, bh->b_size);
 			jh->b_committed_data = NULL;
 			jbd_unlock_bh_state(bh);
 		}
 		jbd2_journal_refile_buffer(journal, jh);
 	}
 	/*
 	 * Now try to drop any written-back buffers from the journal's
 	 * checkpoint lists.  We do this *before* commit because it potentially
 	 * frees some memory
 	 */
 	spin_lock(&journal->j_list_lock);
 	__jbd2_journal_clean_checkpoint_list(journal);
 	spin_unlock(&journal->j_list_lock);
 	jbd_debug (3, "JBD: commit phase 1\n");
 	/*
 	 * Switch to a new revoke table.
 	 */
 	jbd2_journal_switch_revoke_table(journal);
 	trace_jbd2_commit_flushing(journal, commit_transaction);
 	stats.run.rs_flushing = jiffies;
 	stats.run.rs_locked = jbd2_time_diff(stats.run.rs_locked,
 					     stats.run.rs_flushing);
 	commit_transaction->t_state = T_FLUSH;
 	journal->j_committing_transaction = commit_transaction;
 	journal->j_running_transaction = NULL;
 	start_time = ktime_get();
 	commit_transaction->t_log_start = journal->j_head;
 	wake_up(&journal->j_wait_transaction_locked);
 	write_unlock(&journal->j_state_lock);
 	jbd_debug (3, "JBD: commit phase 2\n");
 	/*
 	 * Now start flushing things to disk, in the order they appear
 	 * on the transaction lists.  Data blocks go first.
 	 */
 	err = journal_submit_data_buffers(journal, commit_transaction);
 	if (err)
 		jbd2_journal_abort(journal, err);
 	jbd2_journal_write_revoke_records(journal, commit_transaction,
 					  write_op);
 	jbd_debug(3, "JBD: commit phase 2\n");
 	/*
 	 * Way to go: we have now written out all of the data for a
 	 * transaction!  Now comes the tricky part: we need to write out
 	 * metadata.  Loop over the transaction's entire buffer list:
 	 */
 	write_lock(&journal->j_state_lock);
 	commit_transaction->t_state = T_COMMIT;
 	write_unlock(&journal->j_state_lock);
 	trace_jbd2_commit_logging(journal, commit_transaction);
 	stats.run.rs_logging = jiffies;
 	stats.run.rs_flushing = jbd2_time_diff(stats.run.rs_flushing,
 					       stats.run.rs_logging);
 	stats.run.rs_blocks =
 		atomic_read(&commit_transaction->t_outstanding_credits);
 	stats.run.rs_blocks_logged = 0;
 	J_ASSERT(commit_transaction->t_nr_buffers <=
 		 atomic_read(&commit_transaction->t_outstanding_credits));
 	err = 0;
 	descriptor = NULL;
 	bufs = 0;
 	while (commit_transaction->t_buffers) {
 		/* Find the next buffer to be journaled... */
 		jh = commit_transaction->t_buffers;
 		/* If we're in abort mode, we just un-journal the buffer and
 		   release it. */
 		if (is_journal_aborted(journal)) {
 			clear_buffer_jbddirty(jh2bh(jh));
 			JBUFFER_TRACE(jh, "journal is aborting: refile");
 			jbd2_buffer_abort_trigger(jh,
 						  jh->b_frozen_data ?
 						  jh->b_frozen_triggers :
 						  jh->b_triggers);
 			jbd2_journal_refile_buffer(journal, jh);
 			/* If that was the last one, we need to clean up
 			 * any descriptor buffers which may have been
 			 * already allocated, even if we are now
 			 * aborting. */
 			if (!commit_transaction->t_buffers)
 				goto start_journal_io;
 			continue;
 		}
 		/* Make sure we have a descriptor block in which to
 		   record the metadata buffer. */
 		if (!descriptor) {
 			struct buffer_head *bh;
 			J_ASSERT (bufs == 0);
 			jbd_debug(4, "JBD: get descriptor\n");
 			descriptor = jbd2_journal_get_descriptor_buffer(journal);
 			if (!descriptor) {
 				jbd2_journal_abort(journal, -EIO);
 				continue;
 			}
 			bh = jh2bh(descriptor);
 			jbd_debug(4, "JBD: got buffer %llu (%p)\n",
 				(unsigned long long)bh->b_blocknr, bh->b_data);
 			header = (journal_header_t *)&bh->b_data[0];
 			header->h_magic     = cpu_to_be32(JBD2_MAGIC_NUMBER);
 			header->h_blocktype = cpu_to_be32(JBD2_DESCRIPTOR_BLOCK);
 			header->h_sequence  = cpu_to_be32(commit_transaction->t_tid);
 			tagp = &bh->b_data[sizeof(journal_header_t)];
 			space_left = bh->b_size - sizeof(journal_header_t);
 			first_tag = 1;
 			set_buffer_jwrite(bh);
 			set_buffer_dirty(bh);
 			wbuf[bufs++] = bh;
 			/* Record it so that we can wait for IO
                            completion later */
 			BUFFER_TRACE(bh, "ph3: file as descriptor");
 			jbd2_journal_file_buffer(descriptor, commit_transaction,
 					BJ_LogCtl);
 		}
 		/* Where is the buffer to be written? */
 		err = jbd2_journal_next_log_block(journal, &blocknr);
 		/* If the block mapping failed, just abandon the buffer
 		   and repeat this loop: we'll fall into the
 		   refile-on-abort condition above. */
 		if (err) {
 			jbd2_journal_abort(journal, err);
 			continue;
 		}
 		/*
 		 * start_this_handle() uses t_outstanding_credits to determine
 		 * the free space in the log, but this counter is changed
 		 * by jbd2_journal_next_log_block() also.
 		 */
 		atomic_dec(&commit_transaction->t_outstanding_credits);
 		/* Bump b_count to prevent truncate from stumbling over
                    the shadowed buffer!  @@@ This can go if we ever get
                    rid of the BJ_IO/BJ_Shadow pairing of buffers. */
 		atomic_inc(&jh2bh(jh)->b_count);
 		/* Make a temporary IO buffer with which to write it out
                    (this will requeue both the metadata buffer and the
                    temporary IO buffer). new_bh goes on BJ_IO*/
 		set_bit(BH_JWrite, &jh2bh(jh)->b_state);
 		/*
 		 * akpm: jbd2_journal_write_metadata_buffer() sets
 		 * new_bh->b_transaction to commit_transaction.
 		 * We need to clean this up before we release new_bh
 		 * (which is of type BJ_IO)
 		 */
 		JBUFFER_TRACE(jh, "ph3: write metadata");
 		flags = jbd2_journal_write_metadata_buffer(commit_transaction,
 						      jh, &new_jh, blocknr);
 		if (flags < 0) {
 			jbd2_journal_abort(journal, flags);
 			continue;
 		}
 		set_bit(BH_JWrite, &jh2bh(new_jh)->b_state);
 		wbuf[bufs++] = jh2bh(new_jh);
 		/* Record the new block's tag in the current descriptor
                    buffer */
 		tag_flag = 0;
 		if (flags & 1)
 			tag_flag |= JBD2_FLAG_ESCAPE;
 		if (!first_tag)
 			tag_flag |= JBD2_FLAG_SAME_UUID;
 		tag = (journal_block_tag_t *) tagp;
 		write_tag_block(tag_bytes, tag, jh2bh(jh)->b_blocknr);
 		tag->t_flags = cpu_to_be32(tag_flag);
 		tagp += tag_bytes;
 		space_left -= tag_bytes;
 		if (first_tag) {
 			memcpy (tagp, journal->j_uuid, 16);
 			tagp += 16;
 			space_left -= 16;
 			first_tag = 0;
 		}
 		/* If there's no more to do, or if the descriptor is full,
 		   let the IO rip! */
 		if (bufs == journal->j_wbufsize ||
 		    commit_transaction->t_buffers == NULL ||
 		    space_left < tag_bytes + 16) {
 			jbd_debug(4, "JBD: Submit %d IOs\n", bufs);
 			/* Write an end-of-descriptor marker before
                            submitting the IOs.  "tag" still points to
                            the last tag we set up. */
 			tag->t_flags |= cpu_to_be32(JBD2_FLAG_LAST_TAG);
 start_journal_io:
 			for (i = 0; i < bufs; i++) {
 				struct buffer_head *bh = wbuf[i];
 				/*
 				 * Compute checksum.
 				 */
 				if (JBD2_HAS_COMPAT_FEATURE(journal,
 					JBD2_FEATURE_COMPAT_CHECKSUM)) {
 					crc32_sum =
 					    jbd2_checksum_data(crc32_sum, bh);
 				}
 				lock_buffer(bh);
 				clear_buffer_dirty(bh);
 				set_buffer_uptodate(bh);
 				bh->b_end_io = journal_end_buffer_io_sync;
 				submit_bh(write_op, bh);
 			}
 			cond_resched();
 			stats.run.rs_blocks_logged += bufs;
 			/* Force a new descriptor to be generated next
                            time round the loop. */
 			descriptor = NULL;
 			bufs = 0;
 		}
 	}
 	/*
 	 * If the journal is not located on the file system device,
 	 * then we must flush the file system device before we issue
 	 * the commit record
 	 */
 	if (commit_transaction->t_flushed_data_blocks &&
 	    (journal->j_fs_dev != journal->j_dev) &&
 	    (journal->j_flags & JBD2_BARRIER))
 		blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL,
 			BLKDEV_IFL_WAIT);
 	/* Done it all: now write the commit record asynchronously. */
 	if (JBD2_HAS_INCOMPAT_FEATURE(journal,
 				      JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) {
 		err = journal_submit_commit_record(journal, commit_transaction,
 						 &cbh, crc32_sum);
 		if (err)
 			__jbd2_journal_abort_hard(journal);
 		if (journal->j_flags & JBD2_BARRIER)
 			blkdev_issue_flush(journal->j_dev, GFP_KERNEL, NULL,
 				BLKDEV_IFL_WAIT);
 	}
 	err = journal_finish_inode_data_buffers(journal, commit_transaction);
 	if (err) {
 		printk(KERN_WARNING
 			"JBD2: Detected IO errors while flushing file data "
 		       "on %s\n", journal->j_devname);
 		if (journal->j_flags & JBD2_ABORT_ON_SYNCDATA_ERR)
 			jbd2_journal_abort(journal, err);
 		err = 0;
 	}
 	/* Lo and behold: we have just managed to send a transaction to
            the log.  Before we can commit it, wait for the IO so far to
            complete.  Control buffers being written are on the
            transaction's t_log_list queue, and metadata buffers are on
            the t_iobuf_list queue.
 	   Wait for the buffers in reverse order.  That way we are
 	   less likely to be woken up until all IOs have completed, and
 	   so we incur less scheduling load.
 	*/
 	jbd_debug(3, "JBD: commit phase 3\n");
 	/*
 	 * akpm: these are BJ_IO, and j_list_lock is not needed.
 	 * See __journal_try_to_free_buffer.
 	 */
 wait_for_iobuf:
 	while (commit_transaction->t_iobuf_list != NULL) {
 		struct buffer_head *bh;
 		jh = commit_transaction->t_iobuf_list->b_tprev;
 		bh = jh2bh(jh);
 		if (buffer_locked(bh)) {
 			wait_on_buffer(bh);
 			goto wait_for_iobuf;
 		}
 		if (cond_resched())
 			goto wait_for_iobuf;
 		if (unlikely(!buffer_uptodate(bh)))
 			err = -EIO;
 		clear_buffer_jwrite(bh);
 		JBUFFER_TRACE(jh, "ph4: unfile after journal write");
 		jbd2_journal_unfile_buffer(journal, jh);
 		/*
 		 * ->t_iobuf_list should contain only dummy buffer_heads
 		 * which were created by jbd2_journal_write_metadata_buffer().
 		 */
 		BUFFER_TRACE(bh, "dumping temporary bh");
 		jbd2_journal_put_journal_head(jh);
 		__brelse(bh);
 		J_ASSERT_BH(bh, atomic_read(&bh->b_count) == 0);
 		free_buffer_head(bh);
 		/* We also have to unlock and free the corresponding
                    shadowed buffer */
 		jh = commit_transaction->t_shadow_list->b_tprev;
 		bh = jh2bh(jh);
 		clear_bit(BH_JWrite, &bh->b_state);
 		J_ASSERT_BH(bh, buffer_jbddirty(bh));
 		/* The metadata is now released for reuse, but we need
                    to remember it against this transaction so that when
                    we finally commit, we can do any checkpointing
                    required. */
 		JBUFFER_TRACE(jh, "file as BJ_Forget");
 		jbd2_journal_file_buffer(jh, commit_transaction, BJ_Forget);
 		/* Wake up any transactions which were waiting for this
 		   IO to complete */
 		wake_up_bit(&bh->b_state, BH_Unshadow);
 		JBUFFER_TRACE(jh, "brelse shadowed buffer");
 		__brelse(bh);
 	}
 	J_ASSERT (commit_transaction->t_shadow_list == NULL);
 	jbd_debug(3, "JBD: commit phase 4\n");
 	/* Here we wait for the revoke record and descriptor record buffers */
  wait_for_ctlbuf:
 	while (commit_transaction->t_log_list != NULL) {
 		struct buffer_head *bh;
 		jh = commit_transaction->t_log_list->b_tprev;
 		bh = jh2bh(jh);
 		if (buffer_locked(bh)) {
 			wait_on_buffer(bh);
 			goto wait_for_ctlbuf;
 		}
 		if (cond_resched())
 			goto wait_for_ctlbuf;
 		if (unlikely(!buffer_uptodate(bh)))
 			err = -EIO;
 		BUFFER_TRACE(bh, "ph5: control buffer writeout done: unfile");
 		clear_buffer_jwrite(bh);
 		jbd2_journal_unfile_buffer(journal, jh);
 		jbd2_journal_put_journal_head(jh);
 		__brelse(bh);		/* One for getblk */
 		/* AKPM: bforget here */
 	}
 	if (err)
 		jbd2_journal_abort(journal, err);
 	jbd_debug(3, "JBD: commit phase 5\n");
 	if (!JBD2_HAS_INCOMPAT_FEATURE(journal,
 				       JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) {
 		err = journal_submit_commit_record(journal, commit_transaction,
 						&cbh, crc32_sum);
 		if (err)
 			__jbd2_journal_abort_hard(journal);
 	}
 	if (!err && !is_journal_aborted(journal))
 		err = journal_wait_on_commit_record(journal, cbh);
 	if (err)
 		jbd2_journal_abort(journal, err);
 	/* End of a transaction!  Finally, we can do checkpoint
            processing: any buffers committed as a result of this
            transaction can be removed from any checkpoint list it was on
            before. */
 	jbd_debug(3, "JBD: commit phase 6\n");
 	J_ASSERT(list_empty(&commit_transaction->t_inode_list));
 	J_ASSERT(commit_transaction->t_buffers == NULL);
 	J_ASSERT(commit_transaction->t_checkpoint_list == NULL);
 	J_ASSERT(commit_transaction->t_iobuf_list == NULL);
 	J_ASSERT(commit_transaction->t_shadow_list == NULL);
 	J_ASSERT(commit_transaction->t_log_list == NULL);
 restart_loop:
 	/*
 	 * As there are other places (journal_unmap_buffer()) adding buffers
 	 * to this list we have to be careful and hold the j_list_lock.
 	 */
 	spin_lock(&journal->j_list_lock);
 	while (commit_transaction->t_forget) {
 		transaction_t *cp_transaction;
 		struct buffer_head *bh;
 		jh = commit_transaction->t_forget;
 		spin_unlock(&journal->j_list_lock);
 		bh = jh2bh(jh);
 		jbd_lock_bh_state(bh);
 		J_ASSERT_JH(jh,	jh->b_transaction == commit_transaction);
 		/*
 		 * If there is undo-protected committed data against
 		 * this buffer, then we can remove it now.  If it is a
 		 * buffer needing such protection, the old frozen_data
 		 * field now points to a committed version of the
 		 * buffer, so rotate that field to the new committed
 		 * data.
 		 *
 		 * Otherwise, we can just throw away the frozen data now.
 		 *
 		 * We also know that the frozen data has already fired
 		 * its triggers if they exist, so we can clear that too.
 		 */
 		if (jh->b_committed_data) {
 			jbd2_free(jh->b_committed_data, bh->b_size);
 			jh->b_committed_data = NULL;
 			if (jh->b_frozen_data) {
 				jh->b_committed_data = jh->b_frozen_data;
 				jh->b_frozen_data = NULL;
 				jh->b_frozen_triggers = NULL;
 			}
 		} else if (jh->b_frozen_data) {
 			jbd2_free(jh->b_frozen_data, bh->b_size);
 			jh->b_frozen_data = NULL;
 			jh->b_frozen_triggers = NULL;
 		}
 		spin_lock(&journal->j_list_lock);
 		cp_transaction = jh->b_cp_transaction;
 		if (cp_transaction) {
 			JBUFFER_TRACE(jh, "remove from old cp transaction");
 			cp_transaction->t_chp_stats.cs_dropped++;
 			__jbd2_journal_remove_checkpoint(jh);
 		}
 		/* Only re-checkpoint the buffer_head if it is marked
 		 * dirty.  If the buffer was added to the BJ_Forget list
 		 * by jbd2_journal_forget, it may no longer be dirty and
 		 * there's no point in keeping a checkpoint record for
 		 * it. */
 		/* A buffer which has been freed while still being
 		 * journaled by a previous transaction may end up still
 		 * being dirty here, but we want to avoid writing back
 		 * that buffer in the future after the "add to orphan"
 		 * operation been committed,  That's not only a performance
 		 * gain, it also stops aliasing problems if the buffer is
 		 * left behind for writeback and gets reallocated for another
 		 * use in a different page. */
 		if (buffer_freed(bh) && !jh->b_next_transaction) {
 			clear_buffer_freed(bh);
 			clear_buffer_jbddirty(bh);
 		}
 		if (buffer_jbddirty(bh)) {
 			JBUFFER_TRACE(jh, "add to new checkpointing trans");
 			__jbd2_journal_insert_checkpoint(jh, commit_transaction);
 			if (is_journal_aborted(journal))
 				clear_buffer_jbddirty(bh);
 			JBUFFER_TRACE(jh, "refile for checkpoint writeback");
 			__jbd2_journal_refile_buffer(jh);
 			jbd_unlock_bh_state(bh);
 		} else {
 			J_ASSERT_BH(bh, !buffer_dirty(bh));
 			/* The buffer on BJ_Forget list and not jbddirty means
 			 * it has been freed by this transaction and hence it
 			 * could not have been reallocated until this
 			 * transaction has committed. *BUT* it could be
 			 * reallocated once we have written all the data to
 			 * disk and before we process the buffer on BJ_Forget
 			 * list. */
 			JBUFFER_TRACE(jh, "refile or unfile freed buffer");
 			__jbd2_journal_refile_buffer(jh);
 			if (!jh->b_transaction) {
 				jbd_unlock_bh_state(bh);
 				 /* needs a brelse */
 				jbd2_journal_remove_journal_head(bh);
 				release_buffer_page(bh);
 			} else
 				jbd_unlock_bh_state(bh);
 		}
 		cond_resched_lock(&journal->j_list_lock);
 	}
 	spin_unlock(&journal->j_list_lock);
 	/*
 	 * This is a bit sleazy.  We use j_list_lock to protect transition
 	 * of a transaction into T_FINISHED state and calling
 	 * __jbd2_journal_drop_transaction(). Otherwise we could race with
 	 * other checkpointing code processing the transaction...
 	 */
 	write_lock(&journal->j_state_lock);
 	spin_lock(&journal->j_list_lock);
 	/*
 	 * Now recheck if some buffers did not get attached to the transaction
 	 * while the lock was dropped...
 	 */
 	if (commit_transaction->t_forget) {
 		spin_unlock(&journal->j_list_lock);
 		write_unlock(&journal->j_state_lock);
 		goto restart_loop;
 	}
 	/* Done with this transaction! */
 	jbd_debug(3, "JBD: commit phase 7\n");
 	J_ASSERT(commit_transaction->t_state == T_COMMIT);
 	commit_transaction->t_start = jiffies;
 	stats.run.rs_logging = jbd2_time_diff(stats.run.rs_logging,
 					      commit_transaction->t_start);
 	/*
 	 * File the transaction statistics
 	 */
 	stats.ts_tid = commit_transaction->t_tid;
 	stats.run.rs_handle_count =
 		atomic_read(&commit_transaction->t_handle_count);
 	trace_jbd2_run_stats(journal->j_fs_dev->bd_dev,
 			     commit_transaction->t_tid, &stats.run);
 	/*
 	 * Calculate overall stats
 	 */
 	spin_lock(&journal->j_history_lock);
 	journal->j_stats.ts_tid++;
 	journal->j_stats.run.rs_wait += stats.run.rs_wait;
 	journal->j_stats.run.rs_running += stats.run.rs_running;
 	journal->j_stats.run.rs_locked += stats.run.rs_locked;
 	journal->j_stats.run.rs_flushing += stats.run.rs_flushing;
 	journal->j_stats.run.rs_logging += stats.run.rs_logging;
 	journal->j_stats.run.rs_handle_count += stats.run.rs_handle_count;
 	journal->j_stats.run.rs_blocks += stats.run.rs_blocks;
 	journal->j_stats.run.rs_blocks_logged += stats.run.rs_blocks_logged;
 	spin_unlock(&journal->j_history_lock);
 	commit_transaction->t_state = T_FINISHED;
 	J_ASSERT(commit_transaction == journal->j_committing_transaction);
 	journal->j_commit_sequence = commit_transaction->t_tid;
 	journal->j_committing_transaction = NULL;
 	commit_time = ktime_to_ns(ktime_sub(ktime_get(), start_time));
 	/*
 	 * weight the commit time higher than the average time so we don't
 	 * react too strongly to vast changes in the commit time
 	 */
 	if (likely(journal->j_average_commit_time))
 		journal->j_average_commit_time = (commit_time +
 				journal->j_average_commit_time*3) / 4;
 	else
 		journal->j_average_commit_time = commit_time;
 	write_unlock(&journal->j_state_lock);
 	if (commit_transaction->t_checkpoint_list == NULL &&
 	    commit_transaction->t_checkpoint_io_list == NULL) {
 		__jbd2_journal_drop_transaction(journal, commit_transaction);
 		to_free = 1;
 	} else {
 		if (journal->j_checkpoint_transactions == NULL) {
 			journal->j_checkpoint_transactions = commit_transaction;
 			commit_transaction->t_cpnext = commit_transaction;
 			commit_transaction->t_cpprev = commit_transaction;
 		} else {
 			commit_transaction->t_cpnext =
 				journal->j_checkpoint_transactions;
 			commit_transaction->t_cpprev =
 				commit_transaction->t_cpnext->t_cpprev;
 			commit_transaction->t_cpnext->t_cpprev =
 				commit_transaction;
 			commit_transaction->t_cpprev->t_cpnext =
 				commit_transaction;
 		}
 	}
 	spin_unlock(&journal->j_list_lock);
 	if (journal->j_commit_callback)
 		journal->j_commit_callback(journal, commit_transaction);
 	trace_jbd2_end_commit(journal, commit_transaction);
 	jbd_debug(1, "JBD: commit %d complete, head %d\n",
 		  journal->j_commit_sequence, journal->j_tail_sequence);
 	if (to_free)
 		kfree(commit_transaction);
 	wake_up(&journal->j_wait_done_commit);

fs/nilfs2/super.c

Diff comments View file @ 87e9951

1	/*	1	/*
2	* super.c - NILFS module and super block management.	2	* super.c - NILFS module and super block management.
3	*	3	*
4	* Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation.	4	* Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation.
5	*	5	*
6	* This program is free software; you can redistribute it and/or modify	6	* This program is free software; you can redistribute it and/or modify
7	* it under the terms of the GNU General Public License as published by	7	* it under the terms of the GNU General Public License as published by
8	* the Free Software Foundation; either version 2 of the License, or	8	* the Free Software Foundation; either version 2 of the License, or
9	* (at your option) any later version.	9	* (at your option) any later version.
10	*	10	*
11	* This program is distributed in the hope that it will be useful,	11	* This program is distributed in the hope that it will be useful,
12	* but WITHOUT ANY WARRANTY; without even the implied warranty of	12	* but WITHOUT ANY WARRANTY; without even the implied warranty of
13	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the	13	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14	* GNU General Public License for more details.	14	* GNU General Public License for more details.
15	*	15	*
16	* You should have received a copy of the GNU General Public License	16	* You should have received a copy of the GNU General Public License
17	* along with this program; if not, write to the Free Software	17	* along with this program; if not, write to the Free Software
18	* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA	18	* Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
19	*	19	*
20	* Written by Ryusuke Konishi <ryusuke@osrg.net>	20	* Written by Ryusuke Konishi <ryusuke@osrg.net>
21	*/	21	*/
22	/*	22	/*
23	* linux/fs/ext2/super.c	23	* linux/fs/ext2/super.c
24	*	24	*
25	* Copyright (C) 1992, 1993, 1994, 1995	25	* Copyright (C) 1992, 1993, 1994, 1995
26	* Remy Card (card@masi.ibp.fr)	26	* Remy Card (card@masi.ibp.fr)
27	* Laboratoire MASI - Institut Blaise Pascal	27	* Laboratoire MASI - Institut Blaise Pascal
28	* Universite Pierre et Marie Curie (Paris VI)	28	* Universite Pierre et Marie Curie (Paris VI)
29	*	29	*
30	* from	30	* from
31	*	31	*
32	* linux/fs/minix/inode.c	32	* linux/fs/minix/inode.c
33	*	33	*
34	* Copyright (C) 1991, 1992 Linus Torvalds	34	* Copyright (C) 1991, 1992 Linus Torvalds
35	*	35	*
36	* Big-endian to little-endian byte-swapping/bitmaps by	36	* Big-endian to little-endian byte-swapping/bitmaps by
37	* David S. Miller (davem@caip.rutgers.edu), 1995	37	* David S. Miller (davem@caip.rutgers.edu), 1995
38	*/	38	*/
39		39
40	#include <linux/module.h>	40	#include <linux/module.h>
41	#include <linux/string.h>	41	#include <linux/string.h>
42	#include <linux/slab.h>	42	#include <linux/slab.h>
43	#include <linux/init.h>	43	#include <linux/init.h>
44	#include <linux/blkdev.h>	44	#include <linux/blkdev.h>
45	#include <linux/parser.h>	45	#include <linux/parser.h>
46	#include <linux/random.h>	46	#include <linux/random.h>
47	#include <linux/crc32.h>	47	#include <linux/crc32.h>
48	#include <linux/smp_lock.h>	48	#include <linux/smp_lock.h>
49	#include <linux/vfs.h>	49	#include <linux/vfs.h>
50	#include <linux/writeback.h>	50	#include <linux/writeback.h>
51	#include <linux/kobject.h>	51	#include <linux/kobject.h>
52	#include <linux/exportfs.h>	52	#include <linux/exportfs.h>
53	#include <linux/seq_file.h>	53	#include <linux/seq_file.h>
54	#include <linux/mount.h>	54	#include <linux/mount.h>
55	#include "nilfs.h"	55	#include "nilfs.h"
56	#include "mdt.h"	56	#include "mdt.h"
57	#include "alloc.h"	57	#include "alloc.h"
58	#include "btree.h"	58	#include "btree.h"
59	#include "btnode.h"	59	#include "btnode.h"
60	#include "page.h"	60	#include "page.h"
61	#include "cpfile.h"	61	#include "cpfile.h"
62	#include "ifile.h"	62	#include "ifile.h"
63	#include "dat.h"	63	#include "dat.h"
64	#include "segment.h"	64	#include "segment.h"
65	#include "segbuf.h"	65	#include "segbuf.h"
66		66
67	MODULE_AUTHOR("NTT Corp.");	67	MODULE_AUTHOR("NTT Corp.");
68	MODULE_DESCRIPTION("A New Implementation of the Log-structured Filesystem "	68	MODULE_DESCRIPTION("A New Implementation of the Log-structured Filesystem "
69	"(NILFS)");	69	"(NILFS)");
70	MODULE_LICENSE("GPL");	70	MODULE_LICENSE("GPL");
71		71
72	struct kmem_cache *nilfs_inode_cachep;	72	struct kmem_cache *nilfs_inode_cachep;
73	struct kmem_cache *nilfs_transaction_cachep;	73	struct kmem_cache *nilfs_transaction_cachep;
74	struct kmem_cache *nilfs_segbuf_cachep;	74	struct kmem_cache *nilfs_segbuf_cachep;
75	struct kmem_cache *nilfs_btree_path_cache;	75	struct kmem_cache *nilfs_btree_path_cache;
76		76
77	static int nilfs_remount(struct super_block sb, int flags, char *data);	77	static int nilfs_remount(struct super_block sb, int flags, char *data);
78		78
79	static void nilfs_set_error(struct nilfs_sb_info *sbi)	79	static void nilfs_set_error(struct nilfs_sb_info *sbi)
80	{	80	{
81	struct the_nilfs *nilfs = sbi->s_nilfs;	81	struct the_nilfs *nilfs = sbi->s_nilfs;
82	struct nilfs_super_block **sbp;	82	struct nilfs_super_block **sbp;
83		83
84	down_write(&nilfs->ns_sem);	84	down_write(&nilfs->ns_sem);
85	if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) {	85	if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) {
86	nilfs->ns_mount_state \|= NILFS_ERROR_FS;	86	nilfs->ns_mount_state \|= NILFS_ERROR_FS;
87	sbp = nilfs_prepare_super(sbi, 0);	87	sbp = nilfs_prepare_super(sbi, 0);
88	if (likely(sbp)) {	88	if (likely(sbp)) {
89	sbp[0]->s_state \|= cpu_to_le16(NILFS_ERROR_FS);	89	sbp[0]->s_state \|= cpu_to_le16(NILFS_ERROR_FS);
90	if (sbp[1])	90	if (sbp[1])
91	sbp[1]->s_state \|= cpu_to_le16(NILFS_ERROR_FS);	91	sbp[1]->s_state \|= cpu_to_le16(NILFS_ERROR_FS);
92	nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);	92	nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);
93	}	93	}
94	}	94	}
95	up_write(&nilfs->ns_sem);	95	up_write(&nilfs->ns_sem);
96	}	96	}
97		97
98	/**	98	/**
99	* nilfs_error() - report failure condition on a filesystem	99	* nilfs_error() - report failure condition on a filesystem
100	*	100	*
101	* nilfs_error() sets an ERROR_FS flag on the superblock as well as	101	* nilfs_error() sets an ERROR_FS flag on the superblock as well as
102	* reporting an error message. It should be called when NILFS detects	102	* reporting an error message. It should be called when NILFS detects
103	* incoherences or defects of meta data on disk. As for sustainable	103	* incoherences or defects of meta data on disk. As for sustainable
104	* errors such as a single-shot I/O error, nilfs_warning() or the printk()	104	* errors such as a single-shot I/O error, nilfs_warning() or the printk()
105	* function should be used instead.	105	* function should be used instead.
106	*	106	*
107	* The segment constructor must not call this function because it can	107	* The segment constructor must not call this function because it can
108	* kill itself.	108	* kill itself.
109	*/	109	*/
110	void nilfs_error(struct super_block sb, const char function,	110	void nilfs_error(struct super_block sb, const char function,
111	const char *fmt, ...)	111	const char *fmt, ...)
112	{	112	{
113	struct nilfs_sb_info *sbi = NILFS_SB(sb);	113	struct nilfs_sb_info *sbi = NILFS_SB(sb);
114	va_list args;	114	va_list args;
115		115
116	va_start(args, fmt);	116	va_start(args, fmt);
117	printk(KERN_CRIT "NILFS error (device %s): %s: ", sb->s_id, function);	117	printk(KERN_CRIT "NILFS error (device %s): %s: ", sb->s_id, function);
118	vprintk(fmt, args);	118	vprintk(fmt, args);
119	printk("\n");	119	printk("\n");
120	va_end(args);	120	va_end(args);
121		121
122	if (!(sb->s_flags & MS_RDONLY)) {	122	if (!(sb->s_flags & MS_RDONLY)) {
123	nilfs_set_error(sbi);	123	nilfs_set_error(sbi);
124		124
125	if (nilfs_test_opt(sbi, ERRORS_RO)) {	125	if (nilfs_test_opt(sbi, ERRORS_RO)) {
126	printk(KERN_CRIT "Remounting filesystem read-only\n");	126	printk(KERN_CRIT "Remounting filesystem read-only\n");
127	sb->s_flags \|= MS_RDONLY;	127	sb->s_flags \|= MS_RDONLY;
128	}	128	}
129	}	129	}
130		130
131	if (nilfs_test_opt(sbi, ERRORS_PANIC))	131	if (nilfs_test_opt(sbi, ERRORS_PANIC))
132	panic("NILFS (device %s): panic forced after error\n",	132	panic("NILFS (device %s): panic forced after error\n",
133	sb->s_id);	133	sb->s_id);
134	}	134	}
135		135
136	void nilfs_warning(struct super_block sb, const char function,	136	void nilfs_warning(struct super_block sb, const char function,
137	const char *fmt, ...)	137	const char *fmt, ...)
138	{	138	{
139	va_list args;	139	va_list args;
140		140
141	va_start(args, fmt);	141	va_start(args, fmt);
142	printk(KERN_WARNING "NILFS warning (device %s): %s: ",	142	printk(KERN_WARNING "NILFS warning (device %s): %s: ",
143	sb->s_id, function);	143	sb->s_id, function);
144	vprintk(fmt, args);	144	vprintk(fmt, args);
145	printk("\n");	145	printk("\n");
146	va_end(args);	146	va_end(args);
147	}	147	}
148		148
149		149
150	struct inode nilfs_alloc_inode_common(struct the_nilfs nilfs)	150	struct inode nilfs_alloc_inode_common(struct the_nilfs nilfs)
151	{	151	{
152	struct nilfs_inode_info *ii;	152	struct nilfs_inode_info *ii;
153		153
154	ii = kmem_cache_alloc(nilfs_inode_cachep, GFP_NOFS);	154	ii = kmem_cache_alloc(nilfs_inode_cachep, GFP_NOFS);
155	if (!ii)	155	if (!ii)
156	return NULL;	156	return NULL;
157	ii->i_bh = NULL;	157	ii->i_bh = NULL;
158	ii->i_state = 0;	158	ii->i_state = 0;
159	ii->vfs_inode.i_version = 1;	159	ii->vfs_inode.i_version = 1;
160	nilfs_btnode_cache_init(&ii->i_btnode_cache, nilfs->ns_bdi);	160	nilfs_btnode_cache_init(&ii->i_btnode_cache, nilfs->ns_bdi);
161	return &ii->vfs_inode;	161	return &ii->vfs_inode;
162	}	162	}
163		163
164	struct inode nilfs_alloc_inode(struct super_block sb)	164	struct inode nilfs_alloc_inode(struct super_block sb)
165	{	165	{
166	return nilfs_alloc_inode_common(NILFS_SB(sb)->s_nilfs);	166	return nilfs_alloc_inode_common(NILFS_SB(sb)->s_nilfs);
167	}	167	}
168		168
169	void nilfs_destroy_inode(struct inode *inode)	169	void nilfs_destroy_inode(struct inode *inode)
170	{	170	{
171	kmem_cache_free(nilfs_inode_cachep, NILFS_I(inode));	171	kmem_cache_free(nilfs_inode_cachep, NILFS_I(inode));
172	}	172	}
173		173
174	static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag)	174	static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag)
175	{	175	{
176	struct the_nilfs *nilfs = sbi->s_nilfs;	176	struct the_nilfs *nilfs = sbi->s_nilfs;
177	int err;	177	int err;
178	int barrier_done = 0;
179		178
180	if (nilfs_test_opt(sbi, BARRIER)) {
181	set_buffer_ordered(nilfs->ns_sbh[0]);
182	barrier_done = 1;
183	}
184	retry:	179	retry:
185	set_buffer_dirty(nilfs->ns_sbh[0]);	180	set_buffer_dirty(nilfs->ns_sbh[0]);
186	err = sync_dirty_buffer(nilfs->ns_sbh[0]);	181
187	if (err == -EOPNOTSUPP && barrier_done) {	182	if (nilfs_test_opt(sbi, BARRIER)) {
188	nilfs_warning(sbi->s_super, __func__,	183	err = __sync_dirty_buffer(nilfs->ns_sbh[0],
189	"barrier-based sync failed. "	184	WRITE_SYNC \| WRITE_BARRIER);
190	"disabling barriers\n");	185	if (err == -EOPNOTSUPP) {
191	nilfs_clear_opt(sbi, BARRIER);	186	nilfs_warning(sbi->s_super, __func__,
192	barrier_done = 0;	187	"barrier-based sync failed. "
193	clear_buffer_ordered(nilfs->ns_sbh[0]);	188	"disabling barriers\n");
194	goto retry;	189	nilfs_clear_opt(sbi, BARRIER);
		190	goto retry;
		191	}
		192	} else {
		193	err = sync_dirty_buffer(nilfs->ns_sbh[0]);
195	}	194	}
		195
196	if (unlikely(err)) {	196	if (unlikely(err)) {
197	printk(KERN_ERR	197	printk(KERN_ERR
198	"NILFS: unable to write superblock (err=%d)\n", err);	198	"NILFS: unable to write superblock (err=%d)\n", err);
199	if (err == -EIO && nilfs->ns_sbh[1]) {	199	if (err == -EIO && nilfs->ns_sbh[1]) {
200	/*	200	/*
201	* sbp[0] points to newer log than sbp[1],	201	* sbp[0] points to newer log than sbp[1],
202	* so copy sbp[0] to sbp[1] to take over sbp[0].	202	* so copy sbp[0] to sbp[1] to take over sbp[0].
203	*/	203	*/
204	memcpy(nilfs->ns_sbp[1], nilfs->ns_sbp[0],	204	memcpy(nilfs->ns_sbp[1], nilfs->ns_sbp[0],
205	nilfs->ns_sbsize);	205	nilfs->ns_sbsize);
206	nilfs_fall_back_super_block(nilfs);	206	nilfs_fall_back_super_block(nilfs);
207	goto retry;	207	goto retry;
208	}	208	}
209	} else {	209	} else {
210	struct nilfs_super_block *sbp = nilfs->ns_sbp[0];	210	struct nilfs_super_block *sbp = nilfs->ns_sbp[0];
211		211
212	nilfs->ns_sbwcount++;	212	nilfs->ns_sbwcount++;
213		213
214	/*	214	/*
215	* The latest segment becomes trailable from the position	215	* The latest segment becomes trailable from the position
216	* written in superblock.	216	* written in superblock.
217	*/	217	*/
218	clear_nilfs_discontinued(nilfs);	218	clear_nilfs_discontinued(nilfs);
219		219
220	/* update GC protection for recent segments */	220	/* update GC protection for recent segments */
221	if (nilfs->ns_sbh[1]) {	221	if (nilfs->ns_sbh[1]) {
222	if (flag == NILFS_SB_COMMIT_ALL) {	222	if (flag == NILFS_SB_COMMIT_ALL) {
223	set_buffer_dirty(nilfs->ns_sbh[1]);	223	set_buffer_dirty(nilfs->ns_sbh[1]);
224	if (sync_dirty_buffer(nilfs->ns_sbh[1]) < 0)	224	if (sync_dirty_buffer(nilfs->ns_sbh[1]) < 0)
225	goto out;	225	goto out;
226	}	226	}
227	if (le64_to_cpu(nilfs->ns_sbp[1]->s_last_cno) <	227	if (le64_to_cpu(nilfs->ns_sbp[1]->s_last_cno) <
228	le64_to_cpu(nilfs->ns_sbp[0]->s_last_cno))	228	le64_to_cpu(nilfs->ns_sbp[0]->s_last_cno))
229	sbp = nilfs->ns_sbp[1];	229	sbp = nilfs->ns_sbp[1];
230	}	230	}
231		231
232	spin_lock(&nilfs->ns_last_segment_lock);	232	spin_lock(&nilfs->ns_last_segment_lock);
233	nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq);	233	nilfs->ns_prot_seq = le64_to_cpu(sbp->s_last_seq);
234	spin_unlock(&nilfs->ns_last_segment_lock);	234	spin_unlock(&nilfs->ns_last_segment_lock);
235	}	235	}
236	out:	236	out:
237	return err;	237	return err;
238	}	238	}
239		239
240	void nilfs_set_log_cursor(struct nilfs_super_block *sbp,	240	void nilfs_set_log_cursor(struct nilfs_super_block *sbp,
241	struct the_nilfs *nilfs)	241	struct the_nilfs *nilfs)
242	{	242	{
243	sector_t nfreeblocks;	243	sector_t nfreeblocks;
244		244
245	/* nilfs->ns_sem must be locked by the caller. */	245	/* nilfs->ns_sem must be locked by the caller. */
246	nilfs_count_free_blocks(nilfs, &nfreeblocks);	246	nilfs_count_free_blocks(nilfs, &nfreeblocks);
247	sbp->s_free_blocks_count = cpu_to_le64(nfreeblocks);	247	sbp->s_free_blocks_count = cpu_to_le64(nfreeblocks);
248		248
249	spin_lock(&nilfs->ns_last_segment_lock);	249	spin_lock(&nilfs->ns_last_segment_lock);
250	sbp->s_last_seq = cpu_to_le64(nilfs->ns_last_seq);	250	sbp->s_last_seq = cpu_to_le64(nilfs->ns_last_seq);
251	sbp->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg);	251	sbp->s_last_pseg = cpu_to_le64(nilfs->ns_last_pseg);
252	sbp->s_last_cno = cpu_to_le64(nilfs->ns_last_cno);	252	sbp->s_last_cno = cpu_to_le64(nilfs->ns_last_cno);
253	spin_unlock(&nilfs->ns_last_segment_lock);	253	spin_unlock(&nilfs->ns_last_segment_lock);
254	}	254	}
255		255
256	struct nilfs_super_block *nilfs_prepare_super(struct nilfs_sb_info sbi,	256	struct nilfs_super_block *nilfs_prepare_super(struct nilfs_sb_info sbi,
257	int flip)	257	int flip)
258	{	258	{
259	struct the_nilfs *nilfs = sbi->s_nilfs;	259	struct the_nilfs *nilfs = sbi->s_nilfs;
260	struct nilfs_super_block **sbp = nilfs->ns_sbp;	260	struct nilfs_super_block **sbp = nilfs->ns_sbp;
261		261
262	/* nilfs->ns_sem must be locked by the caller. */	262	/* nilfs->ns_sem must be locked by the caller. */
263	if (sbp[0]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {	263	if (sbp[0]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {
264	if (sbp[1] &&	264	if (sbp[1] &&
265	sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) {	265	sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) {
266	memcpy(sbp[0], sbp[1], nilfs->ns_sbsize);	266	memcpy(sbp[0], sbp[1], nilfs->ns_sbsize);
267	} else {	267	} else {
268	printk(KERN_CRIT "NILFS: superblock broke on dev %s\n",	268	printk(KERN_CRIT "NILFS: superblock broke on dev %s\n",
269	sbi->s_super->s_id);	269	sbi->s_super->s_id);
270	return NULL;	270	return NULL;
271	}	271	}
272	} else if (sbp[1] &&	272	} else if (sbp[1] &&
273	sbp[1]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {	273	sbp[1]->s_magic != cpu_to_le16(NILFS_SUPER_MAGIC)) {
274	memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);	274	memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
275	}	275	}
276		276
277	if (flip && sbp[1])	277	if (flip && sbp[1])
278	nilfs_swap_super_block(nilfs);	278	nilfs_swap_super_block(nilfs);
279		279
280	return sbp;	280	return sbp;
281	}	281	}
282		282
283	int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag)	283	int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag)
284	{	284	{
285	struct the_nilfs *nilfs = sbi->s_nilfs;	285	struct the_nilfs *nilfs = sbi->s_nilfs;
286	struct nilfs_super_block **sbp = nilfs->ns_sbp;	286	struct nilfs_super_block **sbp = nilfs->ns_sbp;
287	time_t t;	287	time_t t;
288		288
289	/* nilfs->ns_sem must be locked by the caller. */	289	/* nilfs->ns_sem must be locked by the caller. */
290	t = get_seconds();	290	t = get_seconds();
291	nilfs->ns_sbwtime = t;	291	nilfs->ns_sbwtime = t;
292	sbp[0]->s_wtime = cpu_to_le64(t);	292	sbp[0]->s_wtime = cpu_to_le64(t);
293	sbp[0]->s_sum = 0;	293	sbp[0]->s_sum = 0;
294	sbp[0]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,	294	sbp[0]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,
295	(unsigned char *)sbp[0],	295	(unsigned char *)sbp[0],
296	nilfs->ns_sbsize));	296	nilfs->ns_sbsize));
297	if (flag == NILFS_SB_COMMIT_ALL && sbp[1]) {	297	if (flag == NILFS_SB_COMMIT_ALL && sbp[1]) {
298	sbp[1]->s_wtime = sbp[0]->s_wtime;	298	sbp[1]->s_wtime = sbp[0]->s_wtime;
299	sbp[1]->s_sum = 0;	299	sbp[1]->s_sum = 0;
300	sbp[1]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,	300	sbp[1]->s_sum = cpu_to_le32(crc32_le(nilfs->ns_crc_seed,
301	(unsigned char *)sbp[1],	301	(unsigned char *)sbp[1],
302	nilfs->ns_sbsize));	302	nilfs->ns_sbsize));
303	}	303	}
304	clear_nilfs_sb_dirty(nilfs);	304	clear_nilfs_sb_dirty(nilfs);
305	return nilfs_sync_super(sbi, flag);	305	return nilfs_sync_super(sbi, flag);
306	}	306	}
307		307
308	/**	308	/**
309	* nilfs_cleanup_super() - write filesystem state for cleanup	309	* nilfs_cleanup_super() - write filesystem state for cleanup
310	* @sbi: nilfs_sb_info to be unmounted or degraded to read-only	310	* @sbi: nilfs_sb_info to be unmounted or degraded to read-only
311	*	311	*
312	* This function restores state flags in the on-disk super block.	312	* This function restores state flags in the on-disk super block.
313	* This will set "clean" flag (i.e. NILFS_VALID_FS) unless the	313	* This will set "clean" flag (i.e. NILFS_VALID_FS) unless the
314	* filesystem was not clean previously.	314	* filesystem was not clean previously.
315	*/	315	*/
316	int nilfs_cleanup_super(struct nilfs_sb_info *sbi)	316	int nilfs_cleanup_super(struct nilfs_sb_info *sbi)
317	{	317	{
318	struct nilfs_super_block **sbp;	318	struct nilfs_super_block **sbp;
319	int flag = NILFS_SB_COMMIT;	319	int flag = NILFS_SB_COMMIT;
320	int ret = -EIO;	320	int ret = -EIO;
321		321
322	sbp = nilfs_prepare_super(sbi, 0);	322	sbp = nilfs_prepare_super(sbi, 0);
323	if (sbp) {	323	if (sbp) {
324	sbp[0]->s_state = cpu_to_le16(sbi->s_nilfs->ns_mount_state);	324	sbp[0]->s_state = cpu_to_le16(sbi->s_nilfs->ns_mount_state);
325	nilfs_set_log_cursor(sbp[0], sbi->s_nilfs);	325	nilfs_set_log_cursor(sbp[0], sbi->s_nilfs);
326	if (sbp[1] && sbp[0]->s_last_cno == sbp[1]->s_last_cno) {	326	if (sbp[1] && sbp[0]->s_last_cno == sbp[1]->s_last_cno) {
327	/*	327	/*
328	* make the "clean" flag also to the opposite	328	* make the "clean" flag also to the opposite
329	* super block if both super blocks point to	329	* super block if both super blocks point to
330	* the same checkpoint.	330	* the same checkpoint.
331	*/	331	*/
332	sbp[1]->s_state = sbp[0]->s_state;	332	sbp[1]->s_state = sbp[0]->s_state;
333	flag = NILFS_SB_COMMIT_ALL;	333	flag = NILFS_SB_COMMIT_ALL;
334	}	334	}
335	ret = nilfs_commit_super(sbi, flag);	335	ret = nilfs_commit_super(sbi, flag);
336	}	336	}
337	return ret;	337	return ret;
338	}	338	}
339		339
340	static void nilfs_put_super(struct super_block *sb)	340	static void nilfs_put_super(struct super_block *sb)
341	{	341	{
342	struct nilfs_sb_info *sbi = NILFS_SB(sb);	342	struct nilfs_sb_info *sbi = NILFS_SB(sb);
343	struct the_nilfs *nilfs = sbi->s_nilfs;	343	struct the_nilfs *nilfs = sbi->s_nilfs;
344		344
345	lock_kernel();	345	lock_kernel();
346		346
347	nilfs_detach_segment_constructor(sbi);	347	nilfs_detach_segment_constructor(sbi);
348		348
349	if (!(sb->s_flags & MS_RDONLY)) {	349	if (!(sb->s_flags & MS_RDONLY)) {
350	down_write(&nilfs->ns_sem);	350	down_write(&nilfs->ns_sem);
351	nilfs_cleanup_super(sbi);	351	nilfs_cleanup_super(sbi);
352	up_write(&nilfs->ns_sem);	352	up_write(&nilfs->ns_sem);
353	}	353	}
354	down_write(&nilfs->ns_super_sem);	354	down_write(&nilfs->ns_super_sem);
355	if (nilfs->ns_current == sbi)	355	if (nilfs->ns_current == sbi)
356	nilfs->ns_current = NULL;	356	nilfs->ns_current = NULL;
357	up_write(&nilfs->ns_super_sem);	357	up_write(&nilfs->ns_super_sem);
358		358
359	nilfs_detach_checkpoint(sbi);	359	nilfs_detach_checkpoint(sbi);
360	put_nilfs(sbi->s_nilfs);	360	put_nilfs(sbi->s_nilfs);
361	sbi->s_super = NULL;	361	sbi->s_super = NULL;
362	sb->s_fs_info = NULL;	362	sb->s_fs_info = NULL;
363	nilfs_put_sbinfo(sbi);	363	nilfs_put_sbinfo(sbi);
364		364
365	unlock_kernel();	365	unlock_kernel();
366	}	366	}
367		367
368	static int nilfs_sync_fs(struct super_block *sb, int wait)	368	static int nilfs_sync_fs(struct super_block *sb, int wait)
369	{	369	{
370	struct nilfs_sb_info *sbi = NILFS_SB(sb);	370	struct nilfs_sb_info *sbi = NILFS_SB(sb);
371	struct the_nilfs *nilfs = sbi->s_nilfs;	371	struct the_nilfs *nilfs = sbi->s_nilfs;
372	struct nilfs_super_block **sbp;	372	struct nilfs_super_block **sbp;
373	int err = 0;	373	int err = 0;
374		374
375	/* This function is called when super block should be written back */	375	/* This function is called when super block should be written back */
376	if (wait)	376	if (wait)
377	err = nilfs_construct_segment(sb);	377	err = nilfs_construct_segment(sb);
378		378
379	down_write(&nilfs->ns_sem);	379	down_write(&nilfs->ns_sem);
380	if (nilfs_sb_dirty(nilfs)) {	380	if (nilfs_sb_dirty(nilfs)) {
381	sbp = nilfs_prepare_super(sbi, nilfs_sb_will_flip(nilfs));	381	sbp = nilfs_prepare_super(sbi, nilfs_sb_will_flip(nilfs));
382	if (likely(sbp)) {	382	if (likely(sbp)) {
383	nilfs_set_log_cursor(sbp[0], nilfs);	383	nilfs_set_log_cursor(sbp[0], nilfs);
384	nilfs_commit_super(sbi, NILFS_SB_COMMIT);	384	nilfs_commit_super(sbi, NILFS_SB_COMMIT);
385	}	385	}
386	}	386	}
387	up_write(&nilfs->ns_sem);	387	up_write(&nilfs->ns_sem);
388		388
389	return err;	389	return err;
390	}	390	}
391		391
392	int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno)	392	int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno)
393	{	393	{
394	struct the_nilfs *nilfs = sbi->s_nilfs;	394	struct the_nilfs *nilfs = sbi->s_nilfs;
395	struct nilfs_checkpoint *raw_cp;	395	struct nilfs_checkpoint *raw_cp;
396	struct buffer_head *bh_cp;	396	struct buffer_head *bh_cp;
397	int err;	397	int err;
398		398
399	down_write(&nilfs->ns_super_sem);	399	down_write(&nilfs->ns_super_sem);
400	list_add(&sbi->s_list, &nilfs->ns_supers);	400	list_add(&sbi->s_list, &nilfs->ns_supers);
401	up_write(&nilfs->ns_super_sem);	401	up_write(&nilfs->ns_super_sem);
402		402
403	sbi->s_ifile = nilfs_ifile_new(sbi, nilfs->ns_inode_size);	403	sbi->s_ifile = nilfs_ifile_new(sbi, nilfs->ns_inode_size);
404	if (!sbi->s_ifile)	404	if (!sbi->s_ifile)
405	return -ENOMEM;	405	return -ENOMEM;
406		406
407	down_read(&nilfs->ns_segctor_sem);	407	down_read(&nilfs->ns_segctor_sem);
408	err = nilfs_cpfile_get_checkpoint(nilfs->ns_cpfile, cno, 0, &raw_cp,	408	err = nilfs_cpfile_get_checkpoint(nilfs->ns_cpfile, cno, 0, &raw_cp,
409	&bh_cp);	409	&bh_cp);
410	up_read(&nilfs->ns_segctor_sem);	410	up_read(&nilfs->ns_segctor_sem);
411	if (unlikely(err)) {	411	if (unlikely(err)) {
412	if (err == -ENOENT \|\| err == -EINVAL) {	412	if (err == -ENOENT \|\| err == -EINVAL) {
413	printk(KERN_ERR	413	printk(KERN_ERR
414	"NILFS: Invalid checkpoint "	414	"NILFS: Invalid checkpoint "
415	"(checkpoint number=%llu)\n",	415	"(checkpoint number=%llu)\n",
416	(unsigned long long)cno);	416	(unsigned long long)cno);
417	err = -EINVAL;	417	err = -EINVAL;
418	}	418	}
419	goto failed;	419	goto failed;
420	}	420	}
421	err = nilfs_read_inode_common(sbi->s_ifile, &raw_cp->cp_ifile_inode);	421	err = nilfs_read_inode_common(sbi->s_ifile, &raw_cp->cp_ifile_inode);
422	if (unlikely(err))	422	if (unlikely(err))
423	goto failed_bh;	423	goto failed_bh;
424	atomic_set(&sbi->s_inodes_count, le64_to_cpu(raw_cp->cp_inodes_count));	424	atomic_set(&sbi->s_inodes_count, le64_to_cpu(raw_cp->cp_inodes_count));
425	atomic_set(&sbi->s_blocks_count, le64_to_cpu(raw_cp->cp_blocks_count));	425	atomic_set(&sbi->s_blocks_count, le64_to_cpu(raw_cp->cp_blocks_count));
426		426
427	nilfs_cpfile_put_checkpoint(nilfs->ns_cpfile, cno, bh_cp);	427	nilfs_cpfile_put_checkpoint(nilfs->ns_cpfile, cno, bh_cp);
428	return 0;	428	return 0;
429		429
430	failed_bh:	430	failed_bh:
431	nilfs_cpfile_put_checkpoint(nilfs->ns_cpfile, cno, bh_cp);	431	nilfs_cpfile_put_checkpoint(nilfs->ns_cpfile, cno, bh_cp);
432	failed:	432	failed:
433	nilfs_mdt_destroy(sbi->s_ifile);	433	nilfs_mdt_destroy(sbi->s_ifile);
434	sbi->s_ifile = NULL;	434	sbi->s_ifile = NULL;
435		435
436	down_write(&nilfs->ns_super_sem);	436	down_write(&nilfs->ns_super_sem);
437	list_del_init(&sbi->s_list);	437	list_del_init(&sbi->s_list);
438	up_write(&nilfs->ns_super_sem);	438	up_write(&nilfs->ns_super_sem);
439		439
440	return err;	440	return err;
441	}	441	}
442		442
443	void nilfs_detach_checkpoint(struct nilfs_sb_info *sbi)	443	void nilfs_detach_checkpoint(struct nilfs_sb_info *sbi)
444	{	444	{
445	struct the_nilfs *nilfs = sbi->s_nilfs;	445	struct the_nilfs *nilfs = sbi->s_nilfs;
446		446
447	nilfs_mdt_destroy(sbi->s_ifile);	447	nilfs_mdt_destroy(sbi->s_ifile);
448	sbi->s_ifile = NULL;	448	sbi->s_ifile = NULL;
449	down_write(&nilfs->ns_super_sem);	449	down_write(&nilfs->ns_super_sem);
450	list_del_init(&sbi->s_list);	450	list_del_init(&sbi->s_list);
451	up_write(&nilfs->ns_super_sem);	451	up_write(&nilfs->ns_super_sem);
452	}	452	}
453		453
454	static int nilfs_statfs(struct dentry dentry, struct kstatfs buf)	454	static int nilfs_statfs(struct dentry dentry, struct kstatfs buf)
455	{	455	{
456	struct super_block *sb = dentry->d_sb;	456	struct super_block *sb = dentry->d_sb;
457	struct nilfs_sb_info *sbi = NILFS_SB(sb);	457	struct nilfs_sb_info *sbi = NILFS_SB(sb);
458	struct the_nilfs *nilfs = sbi->s_nilfs;	458	struct the_nilfs *nilfs = sbi->s_nilfs;
459	u64 id = huge_encode_dev(sb->s_bdev->bd_dev);	459	u64 id = huge_encode_dev(sb->s_bdev->bd_dev);
460	unsigned long long blocks;	460	unsigned long long blocks;
461	unsigned long overhead;	461	unsigned long overhead;
462	unsigned long nrsvblocks;	462	unsigned long nrsvblocks;
463	sector_t nfreeblocks;	463	sector_t nfreeblocks;
464	int err;	464	int err;
465		465
466	/*	466	/*
467	* Compute all of the segment blocks	467	* Compute all of the segment blocks
468	*	468	*
469	* The blocks before first segment and after last segment	469	* The blocks before first segment and after last segment
470	* are excluded.	470	* are excluded.
471	*/	471	*/
472	blocks = nilfs->ns_blocks_per_segment * nilfs->ns_nsegments	472	blocks = nilfs->ns_blocks_per_segment * nilfs->ns_nsegments
473	- nilfs->ns_first_data_block;	473	- nilfs->ns_first_data_block;
474	nrsvblocks = nilfs->ns_nrsvsegs * nilfs->ns_blocks_per_segment;	474	nrsvblocks = nilfs->ns_nrsvsegs * nilfs->ns_blocks_per_segment;
475		475
476	/*	476	/*
477	* Compute the overhead	477	* Compute the overhead
478	*	478	*
479	* When distributing meta data blocks outside segment structure,	479	* When distributing meta data blocks outside segment structure,
480	* We must count them as the overhead.	480	* We must count them as the overhead.
481	*/	481	*/
482	overhead = 0;	482	overhead = 0;
483		483
484	err = nilfs_count_free_blocks(nilfs, &nfreeblocks);	484	err = nilfs_count_free_blocks(nilfs, &nfreeblocks);
485	if (unlikely(err))	485	if (unlikely(err))
486	return err;	486	return err;
487		487
488	buf->f_type = NILFS_SUPER_MAGIC;	488	buf->f_type = NILFS_SUPER_MAGIC;
489	buf->f_bsize = sb->s_blocksize;	489	buf->f_bsize = sb->s_blocksize;
490	buf->f_blocks = blocks - overhead;	490	buf->f_blocks = blocks - overhead;
491	buf->f_bfree = nfreeblocks;	491	buf->f_bfree = nfreeblocks;
492	buf->f_bavail = (buf->f_bfree >= nrsvblocks) ?	492	buf->f_bavail = (buf->f_bfree >= nrsvblocks) ?
493	(buf->f_bfree - nrsvblocks) : 0;	493	(buf->f_bfree - nrsvblocks) : 0;
494	buf->f_files = atomic_read(&sbi->s_inodes_count);	494	buf->f_files = atomic_read(&sbi->s_inodes_count);
495	buf->f_ffree = 0; /* nilfs_count_free_inodes(sb); */	495	buf->f_ffree = 0; /* nilfs_count_free_inodes(sb); */
496	buf->f_namelen = NILFS_NAME_LEN;	496	buf->f_namelen = NILFS_NAME_LEN;
497	buf->f_fsid.val[0] = (u32)id;	497	buf->f_fsid.val[0] = (u32)id;
498	buf->f_fsid.val[1] = (u32)(id >> 32);	498	buf->f_fsid.val[1] = (u32)(id >> 32);
499		499
500	return 0;	500	return 0;
501	}	501	}
502		502
503	static int nilfs_show_options(struct seq_file seq, struct vfsmount vfs)	503	static int nilfs_show_options(struct seq_file seq, struct vfsmount vfs)
504	{	504	{
505	struct super_block *sb = vfs->mnt_sb;	505	struct super_block *sb = vfs->mnt_sb;
506	struct nilfs_sb_info *sbi = NILFS_SB(sb);	506	struct nilfs_sb_info *sbi = NILFS_SB(sb);
507		507
508	if (!nilfs_test_opt(sbi, BARRIER))	508	if (!nilfs_test_opt(sbi, BARRIER))
509	seq_puts(seq, ",nobarrier");	509	seq_puts(seq, ",nobarrier");
510	if (nilfs_test_opt(sbi, SNAPSHOT))	510	if (nilfs_test_opt(sbi, SNAPSHOT))
511	seq_printf(seq, ",cp=%llu",	511	seq_printf(seq, ",cp=%llu",
512	(unsigned long long int)sbi->s_snapshot_cno);	512	(unsigned long long int)sbi->s_snapshot_cno);
513	if (nilfs_test_opt(sbi, ERRORS_PANIC))	513	if (nilfs_test_opt(sbi, ERRORS_PANIC))
514	seq_puts(seq, ",errors=panic");	514	seq_puts(seq, ",errors=panic");
515	if (nilfs_test_opt(sbi, ERRORS_CONT))	515	if (nilfs_test_opt(sbi, ERRORS_CONT))
516	seq_puts(seq, ",errors=continue");	516	seq_puts(seq, ",errors=continue");
517	if (nilfs_test_opt(sbi, STRICT_ORDER))	517	if (nilfs_test_opt(sbi, STRICT_ORDER))
518	seq_puts(seq, ",order=strict");	518	seq_puts(seq, ",order=strict");
519	if (nilfs_test_opt(sbi, NORECOVERY))	519	if (nilfs_test_opt(sbi, NORECOVERY))
520	seq_puts(seq, ",norecovery");	520	seq_puts(seq, ",norecovery");
521	if (nilfs_test_opt(sbi, DISCARD))	521	if (nilfs_test_opt(sbi, DISCARD))
522	seq_puts(seq, ",discard");	522	seq_puts(seq, ",discard");
523		523
524	return 0;	524	return 0;
525	}	525	}
526		526
527	static const struct super_operations nilfs_sops = {	527	static const struct super_operations nilfs_sops = {
528	.alloc_inode = nilfs_alloc_inode,	528	.alloc_inode = nilfs_alloc_inode,
529	.destroy_inode = nilfs_destroy_inode,	529	.destroy_inode = nilfs_destroy_inode,
530	.dirty_inode = nilfs_dirty_inode,	530	.dirty_inode = nilfs_dirty_inode,
531	/* .write_inode = nilfs_write_inode, */	531	/* .write_inode = nilfs_write_inode, */
532	/* .put_inode = nilfs_put_inode, */	532	/* .put_inode = nilfs_put_inode, */
533	/* .drop_inode = nilfs_drop_inode, */	533	/* .drop_inode = nilfs_drop_inode, */
534	.evict_inode = nilfs_evict_inode,	534	.evict_inode = nilfs_evict_inode,
535	.put_super = nilfs_put_super,	535	.put_super = nilfs_put_super,
536	/* .write_super = nilfs_write_super, */	536	/* .write_super = nilfs_write_super, */
537	.sync_fs = nilfs_sync_fs,	537	.sync_fs = nilfs_sync_fs,
538	/* .write_super_lockfs */	538	/* .write_super_lockfs */
539	/* .unlockfs */	539	/* .unlockfs */
540	.statfs = nilfs_statfs,	540	.statfs = nilfs_statfs,
541	.remount_fs = nilfs_remount,	541	.remount_fs = nilfs_remount,
542	/* .umount_begin */	542	/* .umount_begin */
543	.show_options = nilfs_show_options	543	.show_options = nilfs_show_options
544	};	544	};
545		545
546	static struct inode *	546	static struct inode *
547	nilfs_nfs_get_inode(struct super_block *sb, u64 ino, u32 generation)	547	nilfs_nfs_get_inode(struct super_block *sb, u64 ino, u32 generation)
548	{	548	{
549	struct inode *inode;	549	struct inode *inode;
550		550
551	if (ino < NILFS_FIRST_INO(sb) && ino != NILFS_ROOT_INO &&	551	if (ino < NILFS_FIRST_INO(sb) && ino != NILFS_ROOT_INO &&
552	ino != NILFS_SKETCH_INO)	552	ino != NILFS_SKETCH_INO)
553	return ERR_PTR(-ESTALE);	553	return ERR_PTR(-ESTALE);
554		554
555	inode = nilfs_iget(sb, ino);	555	inode = nilfs_iget(sb, ino);
556	if (IS_ERR(inode))	556	if (IS_ERR(inode))
557	return ERR_CAST(inode);	557	return ERR_CAST(inode);
558	if (generation && inode->i_generation != generation) {	558	if (generation && inode->i_generation != generation) {
559	iput(inode);	559	iput(inode);
560	return ERR_PTR(-ESTALE);	560	return ERR_PTR(-ESTALE);
561	}	561	}
562		562
563	return inode;	563	return inode;
564	}	564	}
565		565
566	static struct dentry *	566	static struct dentry *
567	nilfs_fh_to_dentry(struct super_block sb, struct fid fid, int fh_len,	567	nilfs_fh_to_dentry(struct super_block sb, struct fid fid, int fh_len,
568	int fh_type)	568	int fh_type)
569	{	569	{
570	return generic_fh_to_dentry(sb, fid, fh_len, fh_type,	570	return generic_fh_to_dentry(sb, fid, fh_len, fh_type,
571	nilfs_nfs_get_inode);	571	nilfs_nfs_get_inode);
572	}	572	}
573		573
574	static struct dentry *	574	static struct dentry *
575	nilfs_fh_to_parent(struct super_block sb, struct fid fid, int fh_len,	575	nilfs_fh_to_parent(struct super_block sb, struct fid fid, int fh_len,
576	int fh_type)	576	int fh_type)
577	{	577	{
578	return generic_fh_to_parent(sb, fid, fh_len, fh_type,	578	return generic_fh_to_parent(sb, fid, fh_len, fh_type,
579	nilfs_nfs_get_inode);	579	nilfs_nfs_get_inode);
580	}	580	}
581		581
582	static const struct export_operations nilfs_export_ops = {	582	static const struct export_operations nilfs_export_ops = {
583	.fh_to_dentry = nilfs_fh_to_dentry,	583	.fh_to_dentry = nilfs_fh_to_dentry,
584	.fh_to_parent = nilfs_fh_to_parent,	584	.fh_to_parent = nilfs_fh_to_parent,
585	.get_parent = nilfs_get_parent,	585	.get_parent = nilfs_get_parent,
586	};	586	};
587		587
588	enum {	588	enum {
589	Opt_err_cont, Opt_err_panic, Opt_err_ro,	589	Opt_err_cont, Opt_err_panic, Opt_err_ro,
590	Opt_barrier, Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery,	590	Opt_barrier, Opt_nobarrier, Opt_snapshot, Opt_order, Opt_norecovery,
591	Opt_discard, Opt_nodiscard, Opt_err,	591	Opt_discard, Opt_nodiscard, Opt_err,
592	};	592	};
593		593
594	static match_table_t tokens = {	594	static match_table_t tokens = {
595	{Opt_err_cont, "errors=continue"},	595	{Opt_err_cont, "errors=continue"},
596	{Opt_err_panic, "errors=panic"},	596	{Opt_err_panic, "errors=panic"},
597	{Opt_err_ro, "errors=remount-ro"},	597	{Opt_err_ro, "errors=remount-ro"},
598	{Opt_barrier, "barrier"},	598	{Opt_barrier, "barrier"},
599	{Opt_nobarrier, "nobarrier"},	599	{Opt_nobarrier, "nobarrier"},
600	{Opt_snapshot, "cp=%u"},	600	{Opt_snapshot, "cp=%u"},
601	{Opt_order, "order=%s"},	601	{Opt_order, "order=%s"},
602	{Opt_norecovery, "norecovery"},	602	{Opt_norecovery, "norecovery"},
603	{Opt_discard, "discard"},	603	{Opt_discard, "discard"},
604	{Opt_nodiscard, "nodiscard"},	604	{Opt_nodiscard, "nodiscard"},
605	{Opt_err, NULL}	605	{Opt_err, NULL}
606	};	606	};
607		607
608	static int parse_options(char options, struct super_block sb, int is_remount)	608	static int parse_options(char options, struct super_block sb, int is_remount)
609	{	609	{
610	struct nilfs_sb_info *sbi = NILFS_SB(sb);	610	struct nilfs_sb_info *sbi = NILFS_SB(sb);
611	char *p;	611	char *p;
612	substring_t args[MAX_OPT_ARGS];	612	substring_t args[MAX_OPT_ARGS];
613	int option;	613	int option;
614		614
615	if (!options)	615	if (!options)
616	return 1;	616	return 1;
617		617
618	while ((p = strsep(&options, ",")) != NULL) {	618	while ((p = strsep(&options, ",")) != NULL) {
619	int token;	619	int token;
620	if (!*p)	620	if (!*p)
621	continue;	621	continue;
622		622
623	token = match_token(p, tokens, args);	623	token = match_token(p, tokens, args);
624	switch (token) {	624	switch (token) {
625	case Opt_barrier:	625	case Opt_barrier:
626	nilfs_set_opt(sbi, BARRIER);	626	nilfs_set_opt(sbi, BARRIER);
627	break;	627	break;
628	case Opt_nobarrier:	628	case Opt_nobarrier:
629	nilfs_clear_opt(sbi, BARRIER);	629	nilfs_clear_opt(sbi, BARRIER);
630	break;	630	break;
631	case Opt_order:	631	case Opt_order:
632	if (strcmp(args[0].from, "relaxed") == 0)	632	if (strcmp(args[0].from, "relaxed") == 0)
633	/* Ordered data semantics */	633	/* Ordered data semantics */
634	nilfs_clear_opt(sbi, STRICT_ORDER);	634	nilfs_clear_opt(sbi, STRICT_ORDER);
635	else if (strcmp(args[0].from, "strict") == 0)	635	else if (strcmp(args[0].from, "strict") == 0)
636	/* Strict in-order semantics */	636	/* Strict in-order semantics */
637	nilfs_set_opt(sbi, STRICT_ORDER);	637	nilfs_set_opt(sbi, STRICT_ORDER);
638	else	638	else
639	return 0;	639	return 0;
640	break;	640	break;
641	case Opt_err_panic:	641	case Opt_err_panic:
642	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_PANIC);	642	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_PANIC);
643	break;	643	break;
644	case Opt_err_ro:	644	case Opt_err_ro:
645	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_RO);	645	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_RO);
646	break;	646	break;
647	case Opt_err_cont:	647	case Opt_err_cont:
648	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_CONT);	648	nilfs_write_opt(sbi, ERROR_MODE, ERRORS_CONT);
649	break;	649	break;
650	case Opt_snapshot:	650	case Opt_snapshot:
651	if (match_int(&args[0], &option) \|\| option <= 0)	651	if (match_int(&args[0], &option) \|\| option <= 0)
652	return 0;	652	return 0;
653	if (is_remount) {	653	if (is_remount) {
654	if (!nilfs_test_opt(sbi, SNAPSHOT)) {	654	if (!nilfs_test_opt(sbi, SNAPSHOT)) {
655	printk(KERN_ERR	655	printk(KERN_ERR
656	"NILFS: cannot change regular "	656	"NILFS: cannot change regular "
657	"mount to snapshot.\n");	657	"mount to snapshot.\n");
658	return 0;	658	return 0;
659	} else if (option != sbi->s_snapshot_cno) {	659	} else if (option != sbi->s_snapshot_cno) {
660	printk(KERN_ERR	660	printk(KERN_ERR
661	"NILFS: cannot remount to a "	661	"NILFS: cannot remount to a "
662	"different snapshot.\n");	662	"different snapshot.\n");
663	return 0;	663	return 0;
664	}	664	}
665	break;	665	break;
666	}	666	}
667	if (!(sb->s_flags & MS_RDONLY)) {	667	if (!(sb->s_flags & MS_RDONLY)) {
668	printk(KERN_ERR "NILFS: cannot mount snapshot "	668	printk(KERN_ERR "NILFS: cannot mount snapshot "
669	"read/write. A read-only option is "	669	"read/write. A read-only option is "
670	"required.\n");	670	"required.\n");
671	return 0;	671	return 0;
672	}	672	}
673	sbi->s_snapshot_cno = option;	673	sbi->s_snapshot_cno = option;
674	nilfs_set_opt(sbi, SNAPSHOT);	674	nilfs_set_opt(sbi, SNAPSHOT);
675	break;	675	break;
676	case Opt_norecovery:	676	case Opt_norecovery:
677	nilfs_set_opt(sbi, NORECOVERY);	677	nilfs_set_opt(sbi, NORECOVERY);
678	break;	678	break;
679	case Opt_discard:	679	case Opt_discard:
680	nilfs_set_opt(sbi, DISCARD);	680	nilfs_set_opt(sbi, DISCARD);
681	break;	681	break;
682	case Opt_nodiscard:	682	case Opt_nodiscard:
683	nilfs_clear_opt(sbi, DISCARD);	683	nilfs_clear_opt(sbi, DISCARD);
684	break;	684	break;
685	default:	685	default:
686	printk(KERN_ERR	686	printk(KERN_ERR
687	"NILFS: Unrecognized mount option \"%s\"\n", p);	687	"NILFS: Unrecognized mount option \"%s\"\n", p);
688	return 0;	688	return 0;
689	}	689	}
690	}	690	}
691	return 1;	691	return 1;
692	}	692	}
693		693
694	static inline void	694	static inline void
695	nilfs_set_default_options(struct nilfs_sb_info *sbi,	695	nilfs_set_default_options(struct nilfs_sb_info *sbi,
696	struct nilfs_super_block *sbp)	696	struct nilfs_super_block *sbp)
697	{	697	{
698	sbi->s_mount_opt =	698	sbi->s_mount_opt =
699	NILFS_MOUNT_ERRORS_RO \| NILFS_MOUNT_BARRIER;	699	NILFS_MOUNT_ERRORS_RO \| NILFS_MOUNT_BARRIER;
700	}	700	}
701		701
702	static int nilfs_setup_super(struct nilfs_sb_info *sbi)	702	static int nilfs_setup_super(struct nilfs_sb_info *sbi)
703	{	703	{
704	struct the_nilfs *nilfs = sbi->s_nilfs;	704	struct the_nilfs *nilfs = sbi->s_nilfs;
705	struct nilfs_super_block **sbp;	705	struct nilfs_super_block **sbp;
706	int max_mnt_count;	706	int max_mnt_count;
707	int mnt_count;	707	int mnt_count;
708		708
709	/* nilfs->ns_sem must be locked by the caller. */	709	/* nilfs->ns_sem must be locked by the caller. */
710	sbp = nilfs_prepare_super(sbi, 0);	710	sbp = nilfs_prepare_super(sbi, 0);
711	if (!sbp)	711	if (!sbp)
712	return -EIO;	712	return -EIO;
713		713
714	max_mnt_count = le16_to_cpu(sbp[0]->s_max_mnt_count);	714	max_mnt_count = le16_to_cpu(sbp[0]->s_max_mnt_count);
715	mnt_count = le16_to_cpu(sbp[0]->s_mnt_count);	715	mnt_count = le16_to_cpu(sbp[0]->s_mnt_count);
716		716
717	if (nilfs->ns_mount_state & NILFS_ERROR_FS) {	717	if (nilfs->ns_mount_state & NILFS_ERROR_FS) {
718	printk(KERN_WARNING	718	printk(KERN_WARNING
719	"NILFS warning: mounting fs with errors\n");	719	"NILFS warning: mounting fs with errors\n");
720	#if 0	720	#if 0
721	} else if (max_mnt_count >= 0 && mnt_count >= max_mnt_count) {	721	} else if (max_mnt_count >= 0 && mnt_count >= max_mnt_count) {
722	printk(KERN_WARNING	722	printk(KERN_WARNING
723	"NILFS warning: maximal mount count reached\n");	723	"NILFS warning: maximal mount count reached\n");
724	#endif	724	#endif
725	}	725	}
726	if (!max_mnt_count)	726	if (!max_mnt_count)
727	sbp[0]->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT);	727	sbp[0]->s_max_mnt_count = cpu_to_le16(NILFS_DFL_MAX_MNT_COUNT);
728		728
729	sbp[0]->s_mnt_count = cpu_to_le16(mnt_count + 1);	729	sbp[0]->s_mnt_count = cpu_to_le16(mnt_count + 1);
730	sbp[0]->s_state =	730	sbp[0]->s_state =
731	cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & ~NILFS_VALID_FS);	731	cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & ~NILFS_VALID_FS);
732	sbp[0]->s_mtime = cpu_to_le64(get_seconds());	732	sbp[0]->s_mtime = cpu_to_le64(get_seconds());
733	/* synchronize sbp[1] with sbp[0] */	733	/* synchronize sbp[1] with sbp[0] */
734	memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);	734	memcpy(sbp[1], sbp[0], nilfs->ns_sbsize);
735	return nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);	735	return nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL);
736	}	736	}
737		737
738	struct nilfs_super_block nilfs_read_super_block(struct super_block sb,	738	struct nilfs_super_block nilfs_read_super_block(struct super_block sb,
739	u64 pos, int blocksize,	739	u64 pos, int blocksize,
740	struct buffer_head **pbh)	740	struct buffer_head **pbh)
741	{	741	{
742	unsigned long long sb_index = pos;	742	unsigned long long sb_index = pos;
743	unsigned long offset;	743	unsigned long offset;
744		744
745	offset = do_div(sb_index, blocksize);	745	offset = do_div(sb_index, blocksize);
746	*pbh = sb_bread(sb, sb_index);	746	*pbh = sb_bread(sb, sb_index);
747	if (!*pbh)	747	if (!*pbh)
748	return NULL;	748	return NULL;
749	return (struct nilfs_super_block )((char )(*pbh)->b_data + offset);	749	return (struct nilfs_super_block )((char )(*pbh)->b_data + offset);
750	}	750	}
751		751
752	int nilfs_store_magic_and_option(struct super_block *sb,	752	int nilfs_store_magic_and_option(struct super_block *sb,
753	struct nilfs_super_block *sbp,	753	struct nilfs_super_block *sbp,
754	char *data)	754	char *data)
755	{	755	{
756	struct nilfs_sb_info *sbi = NILFS_SB(sb);	756	struct nilfs_sb_info *sbi = NILFS_SB(sb);
757		757
758	sb->s_magic = le16_to_cpu(sbp->s_magic);	758	sb->s_magic = le16_to_cpu(sbp->s_magic);
759		759
760	/* FS independent flags */	760	/* FS independent flags */
761	#ifdef NILFS_ATIME_DISABLE	761	#ifdef NILFS_ATIME_DISABLE
762	sb->s_flags \|= MS_NOATIME;	762	sb->s_flags \|= MS_NOATIME;
763	#endif	763	#endif
764		764
765	nilfs_set_default_options(sbi, sbp);	765	nilfs_set_default_options(sbi, sbp);
766		766
767	sbi->s_resuid = le16_to_cpu(sbp->s_def_resuid);	767	sbi->s_resuid = le16_to_cpu(sbp->s_def_resuid);
768	sbi->s_resgid = le16_to_cpu(sbp->s_def_resgid);	768	sbi->s_resgid = le16_to_cpu(sbp->s_def_resgid);
769	sbi->s_interval = le32_to_cpu(sbp->s_c_interval);	769	sbi->s_interval = le32_to_cpu(sbp->s_c_interval);
770	sbi->s_watermark = le32_to_cpu(sbp->s_c_block_max);	770	sbi->s_watermark = le32_to_cpu(sbp->s_c_block_max);
771		771
772	return !parse_options(data, sb, 0) ? -EINVAL : 0 ;	772	return !parse_options(data, sb, 0) ? -EINVAL : 0 ;
773	}	773	}
774		774
775	int nilfs_check_feature_compatibility(struct super_block *sb,	775	int nilfs_check_feature_compatibility(struct super_block *sb,
776	struct nilfs_super_block *sbp)	776	struct nilfs_super_block *sbp)
777	{	777	{
778	__u64 features;	778	__u64 features;
779		779
780	features = le64_to_cpu(sbp->s_feature_incompat) &	780	features = le64_to_cpu(sbp->s_feature_incompat) &
781	~NILFS_FEATURE_INCOMPAT_SUPP;	781	~NILFS_FEATURE_INCOMPAT_SUPP;
782	if (features) {	782	if (features) {
783	printk(KERN_ERR "NILFS: couldn't mount because of unsupported "	783	printk(KERN_ERR "NILFS: couldn't mount because of unsupported "
784	"optional features (%llx)\n",	784	"optional features (%llx)\n",
785	(unsigned long long)features);	785	(unsigned long long)features);
786	return -EINVAL;	786	return -EINVAL;
787	}	787	}
788	features = le64_to_cpu(sbp->s_feature_compat_ro) &	788	features = le64_to_cpu(sbp->s_feature_compat_ro) &
789	~NILFS_FEATURE_COMPAT_RO_SUPP;	789	~NILFS_FEATURE_COMPAT_RO_SUPP;
790	if (!(sb->s_flags & MS_RDONLY) && features) {	790	if (!(sb->s_flags & MS_RDONLY) && features) {
791	printk(KERN_ERR "NILFS: couldn't mount RDWR because of "	791	printk(KERN_ERR "NILFS: couldn't mount RDWR because of "
792	"unsupported optional features (%llx)\n",	792	"unsupported optional features (%llx)\n",
793	(unsigned long long)features);	793	(unsigned long long)features);
794	return -EINVAL;	794	return -EINVAL;
795	}	795	}
796	return 0;	796	return 0;
797	}	797	}
798		798
799	/**	799	/**
800	* nilfs_fill_super() - initialize a super block instance	800	* nilfs_fill_super() - initialize a super block instance
801	* @sb: super_block	801	* @sb: super_block
802	* @data: mount options	802	* @data: mount options
803	* @silent: silent mode flag	803	* @silent: silent mode flag
804	* @nilfs: the_nilfs struct	804	* @nilfs: the_nilfs struct
805	*	805	*
806	* This function is called exclusively by nilfs->ns_mount_mutex.	806	* This function is called exclusively by nilfs->ns_mount_mutex.
807	* So, the recovery process is protected from other simultaneous mounts.	807	* So, the recovery process is protected from other simultaneous mounts.
808	*/	808	*/
809	static int	809	static int
810	nilfs_fill_super(struct super_block sb, void data, int silent,	810	nilfs_fill_super(struct super_block sb, void data, int silent,
811	struct the_nilfs *nilfs)	811	struct the_nilfs *nilfs)
812	{	812	{
813	struct nilfs_sb_info *sbi;	813	struct nilfs_sb_info *sbi;
814	struct inode *root;	814	struct inode *root;
815	__u64 cno;	815	__u64 cno;
816	int err;	816	int err;
817		817
818	sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);	818	sbi = kzalloc(sizeof(*sbi), GFP_KERNEL);
819	if (!sbi)	819	if (!sbi)
820	return -ENOMEM;	820	return -ENOMEM;
821		821
822	sb->s_fs_info = sbi;	822	sb->s_fs_info = sbi;
823		823
824	get_nilfs(nilfs);	824	get_nilfs(nilfs);
825	sbi->s_nilfs = nilfs;	825	sbi->s_nilfs = nilfs;
826	sbi->s_super = sb;	826	sbi->s_super = sb;
827	atomic_set(&sbi->s_count, 1);	827	atomic_set(&sbi->s_count, 1);
828		828
829	err = init_nilfs(nilfs, sbi, (char *)data);	829	err = init_nilfs(nilfs, sbi, (char *)data);
830	if (err)	830	if (err)
831	goto failed_sbi;	831	goto failed_sbi;
832		832
833	spin_lock_init(&sbi->s_inode_lock);	833	spin_lock_init(&sbi->s_inode_lock);
834	INIT_LIST_HEAD(&sbi->s_dirty_files);	834	INIT_LIST_HEAD(&sbi->s_dirty_files);
835	INIT_LIST_HEAD(&sbi->s_list);	835	INIT_LIST_HEAD(&sbi->s_list);
836		836
837	/*	837	/*
838	* Following initialization is overlapped because	838	* Following initialization is overlapped because
839	* nilfs_sb_info structure has been cleared at the beginning.	839	* nilfs_sb_info structure has been cleared at the beginning.
840	* But we reserve them to keep our interest and make ready	840	* But we reserve them to keep our interest and make ready
841	* for the future change.	841	* for the future change.
842	*/	842	*/
843	get_random_bytes(&sbi->s_next_generation,	843	get_random_bytes(&sbi->s_next_generation,
844	sizeof(sbi->s_next_generation));	844	sizeof(sbi->s_next_generation));
845	spin_lock_init(&sbi->s_next_gen_lock);	845	spin_lock_init(&sbi->s_next_gen_lock);
846		846
847	sb->s_op = &nilfs_sops;	847	sb->s_op = &nilfs_sops;
848	sb->s_export_op = &nilfs_export_ops;	848	sb->s_export_op = &nilfs_export_ops;
849	sb->s_root = NULL;	849	sb->s_root = NULL;
850	sb->s_time_gran = 1;	850	sb->s_time_gran = 1;
851	sb->s_bdi = nilfs->ns_bdi;	851	sb->s_bdi = nilfs->ns_bdi;
852		852
853	err = load_nilfs(nilfs, sbi);	853	err = load_nilfs(nilfs, sbi);
854	if (err)	854	if (err)
855	goto failed_sbi;	855	goto failed_sbi;
856		856
857	cno = nilfs_last_cno(nilfs);	857	cno = nilfs_last_cno(nilfs);
858		858
859	if (sb->s_flags & MS_RDONLY) {	859	if (sb->s_flags & MS_RDONLY) {
860	if (nilfs_test_opt(sbi, SNAPSHOT)) {	860	if (nilfs_test_opt(sbi, SNAPSHOT)) {
861	down_read(&nilfs->ns_segctor_sem);	861	down_read(&nilfs->ns_segctor_sem);
862	err = nilfs_cpfile_is_snapshot(nilfs->ns_cpfile,	862	err = nilfs_cpfile_is_snapshot(nilfs->ns_cpfile,
863	sbi->s_snapshot_cno);	863	sbi->s_snapshot_cno);
864	up_read(&nilfs->ns_segctor_sem);	864	up_read(&nilfs->ns_segctor_sem);
865	if (err < 0) {	865	if (err < 0) {
866	if (err == -ENOENT)	866	if (err == -ENOENT)
867	err = -EINVAL;	867	err = -EINVAL;
868	goto failed_sbi;	868	goto failed_sbi;
869	}	869	}
870	if (!err) {	870	if (!err) {
871	printk(KERN_ERR	871	printk(KERN_ERR
872	"NILFS: The specified checkpoint is "	872	"NILFS: The specified checkpoint is "
873	"not a snapshot "	873	"not a snapshot "
874	"(checkpoint number=%llu).\n",	874	"(checkpoint number=%llu).\n",
875	(unsigned long long)sbi->s_snapshot_cno);	875	(unsigned long long)sbi->s_snapshot_cno);
876	err = -EINVAL;	876	err = -EINVAL;
877	goto failed_sbi;	877	goto failed_sbi;
878	}	878	}
879	cno = sbi->s_snapshot_cno;	879	cno = sbi->s_snapshot_cno;
880	}	880	}
881	}	881	}
882		882
883	err = nilfs_attach_checkpoint(sbi, cno);	883	err = nilfs_attach_checkpoint(sbi, cno);
884	if (err) {	884	if (err) {
885	printk(KERN_ERR "NILFS: error loading a checkpoint"	885	printk(KERN_ERR "NILFS: error loading a checkpoint"
886	" (checkpoint number=%llu).\n", (unsigned long long)cno);	886	" (checkpoint number=%llu).\n", (unsigned long long)cno);
887	goto failed_sbi;	887	goto failed_sbi;
888	}	888	}
889		889
890	if (!(sb->s_flags & MS_RDONLY)) {	890	if (!(sb->s_flags & MS_RDONLY)) {
891	err = nilfs_attach_segment_constructor(sbi);	891	err = nilfs_attach_segment_constructor(sbi);
892	if (err)	892	if (err)
893	goto failed_checkpoint;	893	goto failed_checkpoint;
894	}	894	}
895		895
896	root = nilfs_iget(sb, NILFS_ROOT_INO);	896	root = nilfs_iget(sb, NILFS_ROOT_INO);
897	if (IS_ERR(root)) {	897	if (IS_ERR(root)) {
898	printk(KERN_ERR "NILFS: get root inode failed\n");	898	printk(KERN_ERR "NILFS: get root inode failed\n");
899	err = PTR_ERR(root);	899	err = PTR_ERR(root);
900	goto failed_segctor;	900	goto failed_segctor;
901	}	901	}
902	if (!S_ISDIR(root->i_mode) \|\| !root->i_blocks \|\| !root->i_size) {	902	if (!S_ISDIR(root->i_mode) \|\| !root->i_blocks \|\| !root->i_size) {
903	iput(root);	903	iput(root);
904	printk(KERN_ERR "NILFS: corrupt root inode.\n");	904	printk(KERN_ERR "NILFS: corrupt root inode.\n");
905	err = -EINVAL;	905	err = -EINVAL;
906	goto failed_segctor;	906	goto failed_segctor;
907	}	907	}
908	sb->s_root = d_alloc_root(root);	908	sb->s_root = d_alloc_root(root);
909	if (!sb->s_root) {	909	if (!sb->s_root) {
910	iput(root);	910	iput(root);
911	printk(KERN_ERR "NILFS: get root dentry failed\n");	911	printk(KERN_ERR "NILFS: get root dentry failed\n");
912	err = -ENOMEM;	912	err = -ENOMEM;
913	goto failed_segctor;	913	goto failed_segctor;
914	}	914	}
915		915
916	if (!(sb->s_flags & MS_RDONLY)) {	916	if (!(sb->s_flags & MS_RDONLY)) {
917	down_write(&nilfs->ns_sem);	917	down_write(&nilfs->ns_sem);
918	nilfs_setup_super(sbi);	918	nilfs_setup_super(sbi);
919	up_write(&nilfs->ns_sem);	919	up_write(&nilfs->ns_sem);
920	}	920	}
921		921
922	down_write(&nilfs->ns_super_sem);	922	down_write(&nilfs->ns_super_sem);
923	if (!nilfs_test_opt(sbi, SNAPSHOT))	923	if (!nilfs_test_opt(sbi, SNAPSHOT))
924	nilfs->ns_current = sbi;	924	nilfs->ns_current = sbi;
925	up_write(&nilfs->ns_super_sem);	925	up_write(&nilfs->ns_super_sem);
926		926
927	return 0;	927	return 0;
928		928
929	failed_segctor:	929	failed_segctor:
930	nilfs_detach_segment_constructor(sbi);	930	nilfs_detach_segment_constructor(sbi);
931		931
932	failed_checkpoint:	932	failed_checkpoint:
933	nilfs_detach_checkpoint(sbi);	933	nilfs_detach_checkpoint(sbi);
934		934
935	failed_sbi:	935	failed_sbi:
936	put_nilfs(nilfs);	936	put_nilfs(nilfs);
937	sb->s_fs_info = NULL;	937	sb->s_fs_info = NULL;
938	nilfs_put_sbinfo(sbi);	938	nilfs_put_sbinfo(sbi);
939	return err;	939	return err;
940	}	940	}
941		941
942	static int nilfs_remount(struct super_block sb, int flags, char *data)	942	static int nilfs_remount(struct super_block sb, int flags, char *data)
943	{	943	{
944	struct nilfs_sb_info *sbi = NILFS_SB(sb);	944	struct nilfs_sb_info *sbi = NILFS_SB(sb);
945	struct the_nilfs *nilfs = sbi->s_nilfs;	945	struct the_nilfs *nilfs = sbi->s_nilfs;
946	unsigned long old_sb_flags;	946	unsigned long old_sb_flags;
947	struct nilfs_mount_options old_opts;	947	struct nilfs_mount_options old_opts;
948	int was_snapshot, err;	948	int was_snapshot, err;
949		949
950	lock_kernel();	950	lock_kernel();
951		951
952	down_write(&nilfs->ns_super_sem);	952	down_write(&nilfs->ns_super_sem);
953	old_sb_flags = sb->s_flags;	953	old_sb_flags = sb->s_flags;
954	old_opts.mount_opt = sbi->s_mount_opt;	954	old_opts.mount_opt = sbi->s_mount_opt;
955	old_opts.snapshot_cno = sbi->s_snapshot_cno;	955	old_opts.snapshot_cno = sbi->s_snapshot_cno;
956	was_snapshot = nilfs_test_opt(sbi, SNAPSHOT);	956	was_snapshot = nilfs_test_opt(sbi, SNAPSHOT);
957		957
958	if (!parse_options(data, sb, 1)) {	958	if (!parse_options(data, sb, 1)) {
959	err = -EINVAL;	959	err = -EINVAL;
960	goto restore_opts;	960	goto restore_opts;
961	}	961	}
962	sb->s_flags = (sb->s_flags & ~MS_POSIXACL);	962	sb->s_flags = (sb->s_flags & ~MS_POSIXACL);
963		963
964	err = -EINVAL;	964	err = -EINVAL;
965	if (was_snapshot && !(*flags & MS_RDONLY)) {	965	if (was_snapshot && !(*flags & MS_RDONLY)) {
966	printk(KERN_ERR "NILFS (device %s): cannot remount snapshot "	966	printk(KERN_ERR "NILFS (device %s): cannot remount snapshot "
967	"read/write.\n", sb->s_id);	967	"read/write.\n", sb->s_id);
968	goto restore_opts;	968	goto restore_opts;
969	}	969	}
970		970
971	if (!nilfs_valid_fs(nilfs)) {	971	if (!nilfs_valid_fs(nilfs)) {
972	printk(KERN_WARNING "NILFS (device %s): couldn't "	972	printk(KERN_WARNING "NILFS (device %s): couldn't "
973	"remount because the filesystem is in an "	973	"remount because the filesystem is in an "
974	"incomplete recovery state.\n", sb->s_id);	974	"incomplete recovery state.\n", sb->s_id);
975	goto restore_opts;	975	goto restore_opts;
976	}	976	}
977		977
978	if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))	978	if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))
979	goto out;	979	goto out;
980	if (*flags & MS_RDONLY) {	980	if (*flags & MS_RDONLY) {
981	/* Shutting down the segment constructor */	981	/* Shutting down the segment constructor */
982	nilfs_detach_segment_constructor(sbi);	982	nilfs_detach_segment_constructor(sbi);
983	sb->s_flags \|= MS_RDONLY;	983	sb->s_flags \|= MS_RDONLY;
984		984
985	/*	985	/*
986	* Remounting a valid RW partition RDONLY, so set	986	* Remounting a valid RW partition RDONLY, so set
987	* the RDONLY flag and then mark the partition as valid again.	987	* the RDONLY flag and then mark the partition as valid again.
988	*/	988	*/
989	down_write(&nilfs->ns_sem);	989	down_write(&nilfs->ns_sem);
990	nilfs_cleanup_super(sbi);	990	nilfs_cleanup_super(sbi);
991	up_write(&nilfs->ns_sem);	991	up_write(&nilfs->ns_sem);
992	} else {	992	} else {
993	__u64 features;	993	__u64 features;
994		994
995	/*	995	/*
996	* Mounting a RDONLY partition read-write, so reread and	996	* Mounting a RDONLY partition read-write, so reread and
997	* store the current valid flag. (It may have been changed	997	* store the current valid flag. (It may have been changed
998	* by fsck since we originally mounted the partition.)	998	* by fsck since we originally mounted the partition.)
999	*/	999	*/
1000	down_read(&nilfs->ns_sem);	1000	down_read(&nilfs->ns_sem);
1001	features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) &	1001	features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) &
1002	~NILFS_FEATURE_COMPAT_RO_SUPP;	1002	~NILFS_FEATURE_COMPAT_RO_SUPP;
1003	up_read(&nilfs->ns_sem);	1003	up_read(&nilfs->ns_sem);
1004	if (features) {	1004	if (features) {
1005	printk(KERN_WARNING "NILFS (device %s): couldn't "	1005	printk(KERN_WARNING "NILFS (device %s): couldn't "
1006	"remount RDWR because of unsupported optional "	1006	"remount RDWR because of unsupported optional "
1007	"features (%llx)\n",	1007	"features (%llx)\n",
1008	sb->s_id, (unsigned long long)features);	1008	sb->s_id, (unsigned long long)features);
1009	err = -EROFS;	1009	err = -EROFS;
1010	goto restore_opts;	1010	goto restore_opts;
1011	}	1011	}
1012		1012
1013	sb->s_flags &= ~MS_RDONLY;	1013	sb->s_flags &= ~MS_RDONLY;
1014		1014
1015	err = nilfs_attach_segment_constructor(sbi);	1015	err = nilfs_attach_segment_constructor(sbi);
1016	if (err)	1016	if (err)
1017	goto restore_opts;	1017	goto restore_opts;
1018		1018
1019	down_write(&nilfs->ns_sem);	1019	down_write(&nilfs->ns_sem);
1020	nilfs_setup_super(sbi);	1020	nilfs_setup_super(sbi);
1021	up_write(&nilfs->ns_sem);	1021	up_write(&nilfs->ns_sem);
1022	}	1022	}
1023	out:	1023	out:
1024	up_write(&nilfs->ns_super_sem);	1024	up_write(&nilfs->ns_super_sem);
1025	unlock_kernel();	1025	unlock_kernel();
1026	return 0;	1026	return 0;
1027		1027
1028	restore_opts:	1028	restore_opts:
1029	sb->s_flags = old_sb_flags;	1029	sb->s_flags = old_sb_flags;
1030	sbi->s_mount_opt = old_opts.mount_opt;	1030	sbi->s_mount_opt = old_opts.mount_opt;
1031	sbi->s_snapshot_cno = old_opts.snapshot_cno;	1031	sbi->s_snapshot_cno = old_opts.snapshot_cno;
1032	up_write(&nilfs->ns_super_sem);	1032	up_write(&nilfs->ns_super_sem);
1033	unlock_kernel();	1033	unlock_kernel();
1034	return err;	1034	return err;
1035	}	1035	}
1036		1036
1037	struct nilfs_super_data {	1037	struct nilfs_super_data {
1038	struct block_device *bdev;	1038	struct block_device *bdev;
1039	struct nilfs_sb_info *sbi;	1039	struct nilfs_sb_info *sbi;
1040	__u64 cno;	1040	__u64 cno;
1041	int flags;	1041	int flags;
1042	};	1042	};
1043		1043
1044	/**	1044	/**
1045	* nilfs_identify - pre-read mount options needed to identify mount instance	1045	* nilfs_identify - pre-read mount options needed to identify mount instance
1046	* @data: mount options	1046	* @data: mount options
1047	* @sd: nilfs_super_data	1047	* @sd: nilfs_super_data
1048	*/	1048	*/
1049	static int nilfs_identify(char data, struct nilfs_super_data sd)	1049	static int nilfs_identify(char data, struct nilfs_super_data sd)
1050	{	1050	{
1051	char p, options = data;	1051	char p, options = data;
1052	substring_t args[MAX_OPT_ARGS];	1052	substring_t args[MAX_OPT_ARGS];
1053	int option, token;	1053	int option, token;
1054	int ret = 0;	1054	int ret = 0;
1055		1055
1056	do {	1056	do {
1057	p = strsep(&options, ",");	1057	p = strsep(&options, ",");
1058	if (p != NULL && *p) {	1058	if (p != NULL && *p) {
1059	token = match_token(p, tokens, args);	1059	token = match_token(p, tokens, args);
1060	if (token == Opt_snapshot) {	1060	if (token == Opt_snapshot) {
1061	if (!(sd->flags & MS_RDONLY))	1061	if (!(sd->flags & MS_RDONLY))
1062	ret++;	1062	ret++;
1063	else {	1063	else {
1064	ret = match_int(&args[0], &option);	1064	ret = match_int(&args[0], &option);
1065	if (!ret) {	1065	if (!ret) {
1066	if (option > 0)	1066	if (option > 0)
1067	sd->cno = option;	1067	sd->cno = option;
1068	else	1068	else
1069	ret++;	1069	ret++;
1070	}	1070	}
1071	}	1071	}
1072	}	1072	}
1073	if (ret)	1073	if (ret)
1074	printk(KERN_ERR	1074	printk(KERN_ERR
1075	"NILFS: invalid mount option: %s\n", p);	1075	"NILFS: invalid mount option: %s\n", p);
1076	}	1076	}
1077	if (!options)	1077	if (!options)
1078	break;	1078	break;
1079	BUG_ON(options == data);	1079	BUG_ON(options == data);
1080	*(options - 1) = ',';	1080	*(options - 1) = ',';
1081	} while (!ret);	1081	} while (!ret);
1082	return ret;	1082	return ret;
1083	}	1083	}
1084		1084
1085	static int nilfs_set_bdev_super(struct super_block s, void data)	1085	static int nilfs_set_bdev_super(struct super_block s, void data)
1086	{	1086	{
1087	struct nilfs_super_data *sd = data;	1087	struct nilfs_super_data *sd = data;
1088		1088
1089	s->s_bdev = sd->bdev;	1089	s->s_bdev = sd->bdev;
1090	s->s_dev = s->s_bdev->bd_dev;	1090	s->s_dev = s->s_bdev->bd_dev;
1091	return 0;	1091	return 0;
1092	}	1092	}
1093		1093
1094	static int nilfs_test_bdev_super(struct super_block s, void data)	1094	static int nilfs_test_bdev_super(struct super_block s, void data)
1095	{	1095	{
1096	struct nilfs_super_data *sd = data;	1096	struct nilfs_super_data *sd = data;
1097		1097
1098	return sd->sbi && s->s_fs_info == (void *)sd->sbi;	1098	return sd->sbi && s->s_fs_info == (void *)sd->sbi;
1099	}	1099	}
1100		1100
1101	static int	1101	static int
1102	nilfs_get_sb(struct file_system_type *fs_type, int flags,	1102	nilfs_get_sb(struct file_system_type *fs_type, int flags,
1103	const char dev_name, void data, struct vfsmount *mnt)	1103	const char dev_name, void data, struct vfsmount *mnt)
1104	{	1104	{
1105	struct nilfs_super_data sd;	1105	struct nilfs_super_data sd;
1106	struct super_block *s;	1106	struct super_block *s;
1107	fmode_t mode = FMODE_READ;	1107	fmode_t mode = FMODE_READ;
1108	struct the_nilfs *nilfs;	1108	struct the_nilfs *nilfs;
1109	int err, need_to_close = 1;	1109	int err, need_to_close = 1;
1110		1110
1111	if (!(flags & MS_RDONLY))	1111	if (!(flags & MS_RDONLY))
1112	mode \|= FMODE_WRITE;	1112	mode \|= FMODE_WRITE;
1113		1113
1114	sd.bdev = open_bdev_exclusive(dev_name, mode, fs_type);	1114	sd.bdev = open_bdev_exclusive(dev_name, mode, fs_type);
1115	if (IS_ERR(sd.bdev))	1115	if (IS_ERR(sd.bdev))
1116	return PTR_ERR(sd.bdev);	1116	return PTR_ERR(sd.bdev);
1117		1117
1118	/*	1118	/*
1119	* To get mount instance using sget() vfs-routine, NILFS needs	1119	* To get mount instance using sget() vfs-routine, NILFS needs
1120	* much more information than normal filesystems to identify mount	1120	* much more information than normal filesystems to identify mount
1121	* instance. For snapshot mounts, not only a mount type (ro-mount	1121	* instance. For snapshot mounts, not only a mount type (ro-mount
1122	* or rw-mount) but also a checkpoint number is required.	1122	* or rw-mount) but also a checkpoint number is required.
1123	*/	1123	*/
1124	sd.cno = 0;	1124	sd.cno = 0;
1125	sd.flags = flags;	1125	sd.flags = flags;
1126	if (nilfs_identify((char *)data, &sd)) {	1126	if (nilfs_identify((char *)data, &sd)) {
1127	err = -EINVAL;	1127	err = -EINVAL;
1128	goto failed;	1128	goto failed;
1129	}	1129	}
1130		1130
1131	nilfs = find_or_create_nilfs(sd.bdev);	1131	nilfs = find_or_create_nilfs(sd.bdev);
1132	if (!nilfs) {	1132	if (!nilfs) {
1133	err = -ENOMEM;	1133	err = -ENOMEM;
1134	goto failed;	1134	goto failed;
1135	}	1135	}
1136		1136
1137	mutex_lock(&nilfs->ns_mount_mutex);	1137	mutex_lock(&nilfs->ns_mount_mutex);
1138		1138
1139	if (!sd.cno) {	1139	if (!sd.cno) {
1140	/*	1140	/*
1141	* Check if an exclusive mount exists or not.	1141	* Check if an exclusive mount exists or not.
1142	* Snapshot mounts coexist with a current mount	1142	* Snapshot mounts coexist with a current mount
1143	* (i.e. rw-mount or ro-mount), whereas rw-mount and	1143	* (i.e. rw-mount or ro-mount), whereas rw-mount and
1144	* ro-mount are mutually exclusive.	1144	* ro-mount are mutually exclusive.
1145	*/	1145	*/
1146	down_read(&nilfs->ns_super_sem);	1146	down_read(&nilfs->ns_super_sem);
1147	if (nilfs->ns_current &&	1147	if (nilfs->ns_current &&
1148	((nilfs->ns_current->s_super->s_flags ^ flags)	1148	((nilfs->ns_current->s_super->s_flags ^ flags)
1149	& MS_RDONLY)) {	1149	& MS_RDONLY)) {
1150	up_read(&nilfs->ns_super_sem);	1150	up_read(&nilfs->ns_super_sem);
1151	err = -EBUSY;	1151	err = -EBUSY;
1152	goto failed_unlock;	1152	goto failed_unlock;
1153	}	1153	}
1154	up_read(&nilfs->ns_super_sem);	1154	up_read(&nilfs->ns_super_sem);
1155	}	1155	}
1156		1156
1157	/*	1157	/*
1158	* Find existing nilfs_sb_info struct	1158	* Find existing nilfs_sb_info struct
1159	*/	1159	*/
1160	sd.sbi = nilfs_find_sbinfo(nilfs, !(flags & MS_RDONLY), sd.cno);	1160	sd.sbi = nilfs_find_sbinfo(nilfs, !(flags & MS_RDONLY), sd.cno);
1161		1161
1162	/*	1162	/*
1163	* Get super block instance holding the nilfs_sb_info struct.	1163	* Get super block instance holding the nilfs_sb_info struct.
1164	* A new instance is allocated if no existing mount is present or	1164	* A new instance is allocated if no existing mount is present or
1165	* existing instance has been unmounted.	1165	* existing instance has been unmounted.
1166	*/	1166	*/
1167	s = sget(fs_type, nilfs_test_bdev_super, nilfs_set_bdev_super, &sd);	1167	s = sget(fs_type, nilfs_test_bdev_super, nilfs_set_bdev_super, &sd);
1168	if (sd.sbi)	1168	if (sd.sbi)
1169	nilfs_put_sbinfo(sd.sbi);	1169	nilfs_put_sbinfo(sd.sbi);
1170		1170
1171	if (IS_ERR(s)) {	1171	if (IS_ERR(s)) {
1172	err = PTR_ERR(s);	1172	err = PTR_ERR(s);
1173	goto failed_unlock;	1173	goto failed_unlock;
1174	}	1174	}
1175		1175
1176	if (!s->s_root) {	1176	if (!s->s_root) {
1177	char b[BDEVNAME_SIZE];	1177	char b[BDEVNAME_SIZE];
1178		1178
1179	/* New superblock instance created */	1179	/* New superblock instance created */
1180	s->s_flags = flags;	1180	s->s_flags = flags;
1181	s->s_mode = mode;	1181	s->s_mode = mode;
1182	strlcpy(s->s_id, bdevname(sd.bdev, b), sizeof(s->s_id));	1182	strlcpy(s->s_id, bdevname(sd.bdev, b), sizeof(s->s_id));
1183	sb_set_blocksize(s, block_size(sd.bdev));	1183	sb_set_blocksize(s, block_size(sd.bdev));
1184		1184
1185	err = nilfs_fill_super(s, data, flags & MS_SILENT ? 1 : 0,	1185	err = nilfs_fill_super(s, data, flags & MS_SILENT ? 1 : 0,
1186	nilfs);	1186	nilfs);
1187	if (err)	1187	if (err)
1188	goto cancel_new;	1188	goto cancel_new;
1189		1189
1190	s->s_flags \|= MS_ACTIVE;	1190	s->s_flags \|= MS_ACTIVE;
1191	need_to_close = 0;	1191	need_to_close = 0;
1192	}	1192	}
1193		1193
1194	mutex_unlock(&nilfs->ns_mount_mutex);	1194	mutex_unlock(&nilfs->ns_mount_mutex);
1195	put_nilfs(nilfs);	1195	put_nilfs(nilfs);
1196	if (need_to_close)	1196	if (need_to_close)
1197	close_bdev_exclusive(sd.bdev, mode);	1197	close_bdev_exclusive(sd.bdev, mode);
1198	simple_set_mnt(mnt, s);	1198	simple_set_mnt(mnt, s);
1199	return 0;	1199	return 0;
1200		1200
1201	failed_unlock:	1201	failed_unlock:
1202	mutex_unlock(&nilfs->ns_mount_mutex);	1202	mutex_unlock(&nilfs->ns_mount_mutex);
1203	put_nilfs(nilfs);	1203	put_nilfs(nilfs);
1204	failed:	1204	failed:
1205	close_bdev_exclusive(sd.bdev, mode);	1205	close_bdev_exclusive(sd.bdev, mode);
1206		1206
1207	return err;	1207	return err;
1208		1208
1209	cancel_new:	1209	cancel_new:
1210	/* Abandoning the newly allocated superblock */	1210	/* Abandoning the newly allocated superblock */
1211	mutex_unlock(&nilfs->ns_mount_mutex);	1211	mutex_unlock(&nilfs->ns_mount_mutex);
1212	put_nilfs(nilfs);	1212	put_nilfs(nilfs);
1213	deactivate_locked_super(s);	1213	deactivate_locked_super(s);
1214	/*	1214	/*
1215	* deactivate_locked_super() invokes close_bdev_exclusive().	1215	* deactivate_locked_super() invokes close_bdev_exclusive().
1216	* We must finish all post-cleaning before this call;	1216	* We must finish all post-cleaning before this call;
1217	* put_nilfs() needs the block device.	1217	* put_nilfs() needs the block device.
1218	*/	1218	*/
1219	return err;	1219	return err;
1220	}	1220	}
1221		1221
1222	struct file_system_type nilfs_fs_type = {	1222	struct file_system_type nilfs_fs_type = {
1223	.owner = THIS_MODULE,	1223	.owner = THIS_MODULE,
1224	.name = "nilfs2",	1224	.name = "nilfs2",
1225	.get_sb = nilfs_get_sb,	1225	.get_sb = nilfs_get_sb,
1226	.kill_sb = kill_block_super,	1226	.kill_sb = kill_block_super,
1227	.fs_flags = FS_REQUIRES_DEV,	1227	.fs_flags = FS_REQUIRES_DEV,
1228	};	1228	};
1229		1229
1230	static void nilfs_inode_init_once(void *obj)	1230	static void nilfs_inode_init_once(void *obj)
1231	{	1231	{
1232	struct nilfs_inode_info *ii = obj;	1232	struct nilfs_inode_info *ii = obj;
1233		1233
1234	INIT_LIST_HEAD(&ii->i_dirty);	1234	INIT_LIST_HEAD(&ii->i_dirty);
1235	#ifdef CONFIG_NILFS_XATTR	1235	#ifdef CONFIG_NILFS_XATTR
1236	init_rwsem(&ii->xattr_sem);	1236	init_rwsem(&ii->xattr_sem);
1237	#endif	1237	#endif
1238	nilfs_btnode_cache_init_once(&ii->i_btnode_cache);	1238	nilfs_btnode_cache_init_once(&ii->i_btnode_cache);
1239	ii->i_bmap = &ii->i_bmap_data;	1239	ii->i_bmap = &ii->i_bmap_data;
1240	inode_init_once(&ii->vfs_inode);	1240	inode_init_once(&ii->vfs_inode);
1241	}	1241	}
1242		1242
1243	static void nilfs_segbuf_init_once(void *obj)	1243	static void nilfs_segbuf_init_once(void *obj)
1244	{	1244	{
1245	memset(obj, 0, sizeof(struct nilfs_segment_buffer));	1245	memset(obj, 0, sizeof(struct nilfs_segment_buffer));
1246	}	1246	}
1247		1247
1248	static void nilfs_destroy_cachep(void)	1248	static void nilfs_destroy_cachep(void)
1249	{	1249	{
1250	if (nilfs_inode_cachep)	1250	if (nilfs_inode_cachep)
1251	kmem_cache_destroy(nilfs_inode_cachep);	1251	kmem_cache_destroy(nilfs_inode_cachep);
1252	if (nilfs_transaction_cachep)	1252	if (nilfs_transaction_cachep)
1253	kmem_cache_destroy(nilfs_transaction_cachep);	1253	kmem_cache_destroy(nilfs_transaction_cachep);
1254	if (nilfs_segbuf_cachep)	1254	if (nilfs_segbuf_cachep)
1255	kmem_cache_destroy(nilfs_segbuf_cachep);	1255	kmem_cache_destroy(nilfs_segbuf_cachep);
1256	if (nilfs_btree_path_cache)	1256	if (nilfs_btree_path_cache)
1257	kmem_cache_destroy(nilfs_btree_path_cache);	1257	kmem_cache_destroy(nilfs_btree_path_cache);
1258	}	1258	}
1259		1259
1260	static int __init nilfs_init_cachep(void)	1260	static int __init nilfs_init_cachep(void)
1261	{	1261	{
1262	nilfs_inode_cachep = kmem_cache_create("nilfs2_inode_cache",	1262	nilfs_inode_cachep = kmem_cache_create("nilfs2_inode_cache",
1263	sizeof(struct nilfs_inode_info), 0,	1263	sizeof(struct nilfs_inode_info), 0,
1264	SLAB_RECLAIM_ACCOUNT, nilfs_inode_init_once);	1264	SLAB_RECLAIM_ACCOUNT, nilfs_inode_init_once);
1265	if (!nilfs_inode_cachep)	1265	if (!nilfs_inode_cachep)
1266	goto fail;	1266	goto fail;
1267		1267
1268	nilfs_transaction_cachep = kmem_cache_create("nilfs2_transaction_cache",	1268	nilfs_transaction_cachep = kmem_cache_create("nilfs2_transaction_cache",
1269	sizeof(struct nilfs_transaction_info), 0,	1269	sizeof(struct nilfs_transaction_info), 0,
1270	SLAB_RECLAIM_ACCOUNT, NULL);	1270	SLAB_RECLAIM_ACCOUNT, NULL);
1271	if (!nilfs_transaction_cachep)	1271	if (!nilfs_transaction_cachep)
1272	goto fail;	1272	goto fail;
1273		1273
1274	nilfs_segbuf_cachep = kmem_cache_create("nilfs2_segbuf_cache",	1274	nilfs_segbuf_cachep = kmem_cache_create("nilfs2_segbuf_cache",
1275	sizeof(struct nilfs_segment_buffer), 0,	1275	sizeof(struct nilfs_segment_buffer), 0,
1276	SLAB_RECLAIM_ACCOUNT, nilfs_segbuf_init_once);	1276	SLAB_RECLAIM_ACCOUNT, nilfs_segbuf_init_once);
1277	if (!nilfs_segbuf_cachep)	1277	if (!nilfs_segbuf_cachep)
1278	goto fail;	1278	goto fail;
1279		1279
1280	nilfs_btree_path_cache = kmem_cache_create("nilfs2_btree_path_cache",	1280	nilfs_btree_path_cache = kmem_cache_create("nilfs2_btree_path_cache",
1281	sizeof(struct nilfs_btree_path) * NILFS_BTREE_LEVEL_MAX,	1281	sizeof(struct nilfs_btree_path) * NILFS_BTREE_LEVEL_MAX,
1282	0, 0, NULL);	1282	0, 0, NULL);
1283	if (!nilfs_btree_path_cache)	1283	if (!nilfs_btree_path_cache)
1284	goto fail;	1284	goto fail;
1285		1285
1286	return 0;	1286	return 0;
1287		1287
1288	fail:	1288	fail:
1289	nilfs_destroy_cachep();	1289	nilfs_destroy_cachep();
1290	return -ENOMEM;	1290	return -ENOMEM;
1291	}	1291	}
1292		1292
1293	static int __init init_nilfs_fs(void)	1293	static int __init init_nilfs_fs(void)
1294	{	1294	{
1295	int err;	1295	int err;
1296		1296
1297	err = nilfs_init_cachep();	1297	err = nilfs_init_cachep();
1298	if (err)	1298	if (err)
1299	goto fail;	1299	goto fail;
1300		1300
1301	err = register_filesystem(&nilfs_fs_type);	1301	err = register_filesystem(&nilfs_fs_type);
1302	if (err)	1302	if (err)
1303	goto free_cachep;	1303	goto free_cachep;
1304		1304
1305	printk(KERN_INFO "NILFS version 2 loaded\n");	1305	printk(KERN_INFO "NILFS version 2 loaded\n");
1306	return 0;	1306	return 0;
1307		1307
1308	free_cachep:	1308	free_cachep:
1309	nilfs_destroy_cachep();	1309	nilfs_destroy_cachep();
1310	fail:	1310	fail:
1311	return err;	1311	return err;
1312	}	1312	}
1313		1313
1314	static void __exit exit_nilfs_fs(void)	1314	static void __exit exit_nilfs_fs(void)
1315	{	1315	{
1316	nilfs_destroy_cachep();	1316	nilfs_destroy_cachep();
1317	unregister_filesystem(&nilfs_fs_type);	1317	unregister_filesystem(&nilfs_fs_type);

include/linux/buffer_head.h

Diff comments View file @ 87e9951

 /*
  * include/linux/buffer_head.h
  *
  * Everything to do with buffer_heads.
  */
 #ifndef _LINUX_BUFFER_HEAD_H
 #define _LINUX_BUFFER_HEAD_H
 #include <linux/types.h>
 #include <linux/fs.h>
 #include <linux/linkage.h>
 #include <linux/pagemap.h>
 #include <linux/wait.h>
 #include <asm/atomic.h>
 #ifdef CONFIG_BLOCK
 enum bh_state_bits {
 	BH_Uptodate,	/* Contains valid data */
 	BH_Dirty,	/* Is dirty */
 	BH_Lock,	/* Is locked */
 	BH_Req,		/* Has been submitted for I/O */
 	BH_Uptodate_Lock,/* Used by the first bh in a page, to serialise
 			  * IO completion of other buffers in the page
 			  */
 	BH_Mapped,	/* Has a disk mapping */
 	BH_New,		/* Disk mapping was newly created by get_block */
 	BH_Async_Read,	/* Is under end_buffer_async_read I/O */
 	BH_Async_Write,	/* Is under end_buffer_async_write I/O */
 	BH_Delay,	/* Buffer is not yet allocated on disk */
 	BH_Boundary,	/* Block is followed by a discontiguity */
 	BH_Write_EIO,	/* I/O error on write */
-	BH_Ordered,	/* ordered write */
 	BH_Eopnotsupp,	/* operation not supported (barrier) */
 	BH_Unwritten,	/* Buffer is allocated on disk but not written */
 	BH_Quiet,	/* Buffer Error Prinks to be quiet */
 	BH_PrivateStart,/* not a state bit, but the first bit available
 			 * for private allocation by other entities
 			 */
 };
 #define MAX_BUF_PER_PAGE (PAGE_CACHE_SIZE / 512)
 struct page;
 struct buffer_head;
 struct address_space;
 typedef void (bh_end_io_t)(struct buffer_head *bh, int uptodate);
 /*
  * Historically, a buffer_head was used to map a single block
  * within a page, and of course as the unit of I/O through the
  * filesystem and block layers.  Nowadays the basic I/O unit
  * is the bio, and buffer_heads are used for extracting block
  * mappings (via a get_block_t call), for tracking state within
  * a page (via a page_mapping) and for wrapping bio submission
  * for backward compatibility reasons (e.g. submit_bh).
  */
 struct buffer_head {
 	unsigned long b_state;		/* buffer state bitmap (see above) */
 	struct buffer_head *b_this_page;/* circular list of page's buffers */
 	struct page *b_page;		/* the page this bh is mapped to */
 	sector_t b_blocknr;		/* start block number */
 	size_t b_size;			/* size of mapping */
 	char *b_data;			/* pointer to data within the page */
 	struct block_device *b_bdev;
 	bh_end_io_t *b_end_io;		/* I/O completion */
  	void *b_private;		/* reserved for b_end_io */
 	struct list_head b_assoc_buffers; /* associated with another mapping */
 	struct address_space *b_assoc_map;	/* mapping this buffer is
 						   associated with */
 	atomic_t b_count;		/* users using this buffer_head */
 };
 /*
  * macro tricks to expand the set_buffer_foo(), clear_buffer_foo()
  * and buffer_foo() functions.
  */
 #define BUFFER_FNS(bit, name)						\
 static inline void set_buffer_##name(struct buffer_head *bh)		\
 {									\
 	set_bit(BH_##bit, &(bh)->b_state);				\
 }									\
 static inline void clear_buffer_##name(struct buffer_head *bh)		\
 {									\
 	clear_bit(BH_##bit, &(bh)->b_state);				\
 }									\
 static inline int buffer_##name(const struct buffer_head *bh)		\
 {									\
 	return test_bit(BH_##bit, &(bh)->b_state);			\
 }
 /*
  * test_set_buffer_foo() and test_clear_buffer_foo()
  */
 #define TAS_BUFFER_FNS(bit, name)					\
 static inline int test_set_buffer_##name(struct buffer_head *bh)	\
 {									\
 	return test_and_set_bit(BH_##bit, &(bh)->b_state);		\
 }									\
 static inline int test_clear_buffer_##name(struct buffer_head *bh)	\
 {									\
 	return test_and_clear_bit(BH_##bit, &(bh)->b_state);		\
 }									\
 /*
  * Emit the buffer bitops functions.   Note that there are also functions
  * of the form "mark_buffer_foo()".  These are higher-level functions which
  * do something in addition to setting a b_state bit.
  */
 BUFFER_FNS(Uptodate, uptodate)
 BUFFER_FNS(Dirty, dirty)
 TAS_BUFFER_FNS(Dirty, dirty)
 BUFFER_FNS(Lock, locked)
 BUFFER_FNS(Req, req)
 TAS_BUFFER_FNS(Req, req)
 BUFFER_FNS(Mapped, mapped)
 BUFFER_FNS(New, new)
 BUFFER_FNS(Async_Read, async_read)
 BUFFER_FNS(Async_Write, async_write)
 BUFFER_FNS(Delay, delay)
 BUFFER_FNS(Boundary, boundary)
 BUFFER_FNS(Write_EIO, write_io_error)
-BUFFER_FNS(Ordered, ordered)
 BUFFER_FNS(Eopnotsupp, eopnotsupp)
 BUFFER_FNS(Unwritten, unwritten)
 #define bh_offset(bh)		((unsigned long)(bh)->b_data & ~PAGE_MASK)
 #define touch_buffer(bh)	mark_page_accessed(bh->b_page)
 /* If we *know* page->private refers to buffer_heads */
 #define page_buffers(page)					\
 	({							\
 		BUG_ON(!PagePrivate(page));			\
 		((struct buffer_head *)page_private(page));	\
 	})
 #define page_has_buffers(page)	PagePrivate(page)
 /*
  * Declarations
  */
 void mark_buffer_dirty(struct buffer_head *bh);
 void init_buffer(struct buffer_head *, bh_end_io_t *, void *);
 void set_bh_page(struct buffer_head *bh,
 		struct page *page, unsigned long offset);
 int try_to_free_buffers(struct page *);
 struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size,
 		int retry);
 void create_empty_buffers(struct page *, unsigned long,
 			unsigned long b_state);
 void end_buffer_read_sync(struct buffer_head *bh, int uptodate);
 void end_buffer_write_sync(struct buffer_head *bh, int uptodate);
 void end_buffer_async_write(struct buffer_head *bh, int uptodate);
 /* Things to do with buffers at mapping->private_list */
 void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode);
 int inode_has_buffers(struct inode *);
 void invalidate_inode_buffers(struct inode *);
 int remove_inode_buffers(struct inode *inode);
 int sync_mapping_buffers(struct address_space *mapping);
 void unmap_underlying_metadata(struct block_device *bdev, sector_t block);
 void mark_buffer_async_write(struct buffer_head *bh);
 void __wait_on_buffer(struct buffer_head *);
 wait_queue_head_t *bh_waitq_head(struct buffer_head *bh);
 struct buffer_head *__find_get_block(struct block_device *bdev, sector_t block,
 			unsigned size);
 struct buffer_head *__getblk(struct block_device *bdev, sector_t block,
 			unsigned size);
 void __brelse(struct buffer_head *);
 void __bforget(struct buffer_head *);
 void __breadahead(struct block_device *, sector_t block, unsigned int size);
 struct buffer_head *__bread(struct block_device *, sector_t block, unsigned size);
 void invalidate_bh_lrus(void);
 struct buffer_head *alloc_buffer_head(gfp_t gfp_flags);
 void free_buffer_head(struct buffer_head * bh);
 void unlock_buffer(struct buffer_head *bh);
 void __lock_buffer(struct buffer_head *bh);
 void ll_rw_block(int, int, struct buffer_head * bh[]);
 int sync_dirty_buffer(struct buffer_head *bh);
+int __sync_dirty_buffer(struct buffer_head *bh, int rw);
 int submit_bh(int, struct buffer_head *);
 void write_boundary_block(struct block_device *bdev,
 			sector_t bblock, unsigned blocksize);
 int bh_uptodate_or_lock(struct buffer_head *bh);
 int bh_submit_read(struct buffer_head *bh);
 extern int buffer_heads_over_limit;
 /*
  * Generic address_space_operations implementations for buffer_head-backed
  * address_spaces.
  */
 void block_invalidatepage(struct page *page, unsigned long offset);
 int block_write_full_page(struct page *page, get_block_t *get_block,
 				struct writeback_control *wbc);
 int block_write_full_page_endio(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc, bh_end_io_t *handler);
 int block_read_full_page(struct page*, get_block_t*);
 int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc,
 				unsigned long from);
 int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
 		unsigned flags, struct page **pagep, get_block_t *get_block);
 int __block_write_begin(struct page *page, loff_t pos, unsigned len,
 		get_block_t *get_block);
 int block_write_end(struct file *, struct address_space *,
 				loff_t, unsigned, unsigned,
 				struct page *, void *);
 int generic_write_end(struct file *, struct address_space *,
 				loff_t, unsigned, unsigned,
 				struct page *, void *);
 void page_zero_new_buffers(struct page *page, unsigned from, unsigned to);
 int block_prepare_write(struct page*, unsigned, unsigned, get_block_t*);
 int cont_write_begin(struct file *, struct address_space *, loff_t,
 			unsigned, unsigned, struct page **, void **,
 			get_block_t *, loff_t *);
 int generic_cont_expand_simple(struct inode *inode, loff_t size);
 int block_commit_write(struct page *page, unsigned from, unsigned to);
 int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf,
 				get_block_t get_block);
 void block_sync_page(struct page *);
 sector_t generic_block_bmap(struct address_space *, sector_t, get_block_t *);
 int block_truncate_page(struct address_space *, loff_t, get_block_t *);
 int nobh_write_begin(struct address_space *, loff_t, unsigned, unsigned,
 				struct page **, void **, get_block_t*);
 int nobh_write_end(struct file *, struct address_space *,
 				loff_t, unsigned, unsigned,
 				struct page *, void *);
 int nobh_truncate_page(struct address_space *, loff_t, get_block_t *);
 int nobh_writepage(struct page *page, get_block_t *get_block,
                         struct writeback_control *wbc);
 void buffer_init(void);
 /*
  * inline definitions
  */
 static inline void attach_page_buffers(struct page *page,
 		struct buffer_head *head)
 {
 	page_cache_get(page);
 	SetPagePrivate(page);
 	set_page_private(page, (unsigned long)head);
 }
 static inline void get_bh(struct buffer_head *bh)
 {
         atomic_inc(&bh->b_count);
 }
 static inline void put_bh(struct buffer_head *bh)
 {
         smp_mb__before_atomic_dec();
         atomic_dec(&bh->b_count);
 }
 static inline void brelse(struct buffer_head *bh)
 {
 	if (bh)
 		__brelse(bh);
 }
 static inline void bforget(struct buffer_head *bh)
 {
 	if (bh)
 		__bforget(bh);
 }
 static inline struct buffer_head *
 sb_bread(struct super_block *sb, sector_t block)
 {
 	return __bread(sb->s_bdev, block, sb->s_blocksize);
 }
 static inline void
 sb_breadahead(struct super_block *sb, sector_t block)
 {
 	__breadahead(sb->s_bdev, block, sb->s_blocksize);
 }
 static inline struct buffer_head *
 sb_getblk(struct super_block *sb, sector_t block)
 {
 	return __getblk(sb->s_bdev, block, sb->s_blocksize);
 }
 static inline struct buffer_head *
 sb_find_get_block(struct super_block *sb, sector_t block)
 {
 	return __find_get_block(sb->s_bdev, block, sb->s_blocksize);
 }
 static inline void
 map_bh(struct buffer_head *bh, struct super_block *sb, sector_t block)
 {
 	set_buffer_mapped(bh);
 	bh->b_bdev = sb->s_bdev;
 	bh->b_blocknr = block;
 	bh->b_size = sb->s_blocksize;
 }
 static inline void wait_on_buffer(struct buffer_head *bh)
 {
 	might_sleep();
 	if (buffer_locked(bh))
 		__wait_on_buffer(bh);
 }
 static inline int trylock_buffer(struct buffer_head *bh)
 {
 	return likely(!test_and_set_bit_lock(BH_Lock, &bh->b_state));
 }
 static inline void lock_buffer(struct buffer_head *bh)
 {
 	might_sleep();
 	if (!trylock_buffer(bh))
 		__lock_buffer(bh);
 }
 extern int __set_page_dirty_buffers(struct page *page);
 #else /* CONFIG_BLOCK */
 static inline void buffer_init(void) {}
 static inline int try_to_free_buffers(struct page *page) { return 1; }
 static inline int inode_has_buffers(struct inode *inode) { return 0; }
 static inline void invalidate_inode_buffers(struct inode *inode) {}
 static inline int remove_inode_buffers(struct inode *inode) { return 1; }
 static inline int sync_mapping_buffers(struct address_space *mapping) { return 0; }
 #endif /* CONFIG_BLOCK */
 #endif /* _LINUX_BUFFER_HEAD_H */