inode.c

linux 内核源代码

};

static const struct address_space_operations ext4_writeback_aops = {
	.readpage	= ext4_readpage,
	.readpages	= ext4_readpages,
	.writepage	= ext4_writeback_writepage,
	.sync_page	= block_sync_page,
	.write_begin	= ext4_write_begin,
	.write_end	= ext4_writeback_write_end,
	.bmap		= ext4_bmap,
	.invalidatepage	= ext4_invalidatepage,
	.releasepage	= ext4_releasepage,
	.direct_IO	= ext4_direct_IO,
	.migratepage	= buffer_migrate_page,
};

static const struct address_space_operations ext4_journalled_aops = {
	.readpage	= ext4_readpage,
	.readpages	= ext4_readpages,
	.writepage	= ext4_journalled_writepage,
	.sync_page	= block_sync_page,
	.write_begin	= ext4_write_begin,
	.write_end	= ext4_journalled_write_end,
	.set_page_dirty	= ext4_journalled_set_page_dirty,
	.bmap		= ext4_bmap,
	.invalidatepage	= ext4_invalidatepage,
	.releasepage	= ext4_releasepage,
};

void ext4_set_aops(struct inode *inode)
{
	if (ext4_should_order_data(inode))
		inode->i_mapping->a_ops = &ext4_ordered_aops;
	else if (ext4_should_writeback_data(inode))
		inode->i_mapping->a_ops = &ext4_writeback_aops;
	else
		inode->i_mapping->a_ops = &ext4_journalled_aops;
}

/*
 * ext4_block_truncate_page() zeroes out a mapping from file offset `from'
 * up to the end of the block which corresponds to `from'.
 * This required during truncate. We need to physically zero the tail end
 * of that block so it doesn't yield old data if the file is later grown.
 */
int ext4_block_truncate_page(handle_t *handle, struct page *page,
		struct address_space *mapping, loff_t from)
{
	ext4_fsblk_t index = from >> PAGE_CACHE_SHIFT;
	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	unsigned blocksize, iblock, length, pos;
	struct inode *inode = mapping->host;
	struct buffer_head *bh;
	int err = 0;

	blocksize = inode->i_sb->s_blocksize;
	length = blocksize - (offset & (blocksize - 1));
	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);

	/*
	 * For "nobh" option,  we can only work if we don't need to
	 * read-in the page - otherwise we create buffers to do the IO.
	 */
	if (!page_has_buffers(page) && test_opt(inode->i_sb, NOBH) &&
	     ext4_should_writeback_data(inode) && PageUptodate(page)) {
		zero_user_page(page, offset, length, KM_USER0);
		set_page_dirty(page);
		goto unlock;
	}

	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_freed(bh)) {
		BUFFER_TRACE(bh, "freed: skip");
		goto unlock;
	}

	if (!buffer_mapped(bh)) {
		BUFFER_TRACE(bh, "unmapped");
		ext4_get_block(inode, iblock, bh, 0);
		/* unmapped? It's a hole - nothing to do */
		if (!buffer_mapped(bh)) {
			BUFFER_TRACE(bh, "still unmapped");
			goto unlock;
		}
	}

	/* Ok, it's mapped. Make sure it's up-to-date */
	if (PageUptodate(page))
		set_buffer_uptodate(bh);

	if (!buffer_uptodate(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;
	}

	if (ext4_should_journal_data(inode)) {
		BUFFER_TRACE(bh, "get write access");
		err = ext4_journal_get_write_access(handle, bh);
		if (err)
			goto unlock;
	}

	zero_user_page(page, offset, length, KM_USER0);

	BUFFER_TRACE(bh, "zeroed end of block");

	err = 0;
	if (ext4_should_journal_data(inode)) {
		err = ext4_journal_dirty_metadata(handle, bh);
	} else {
		if (ext4_should_order_data(inode))
			err = ext4_journal_dirty_data(handle, bh);
		mark_buffer_dirty(bh);
	}

unlock:
	unlock_page(page);
	page_cache_release(page);
	return err;
}

/*
 * Probably it should be a library function... search for first non-zero word
 * or memcmp with zero_page, whatever is better for particular architecture.
 * Linus?
 */
static inline int all_zeroes(__le32 *p, __le32 *q)
{
	while (p < q)
		if (*p++)
			return 0;
	return 1;
}

/**
 *	ext4_find_shared - find the indirect blocks for partial truncation.
 *	@inode:	  inode in question
 *	@depth:	  depth of the affected branch
 *	@offsets: offsets of pointers in that branch (see ext4_block_to_path)
 *	@chain:	  place to store the pointers to partial indirect blocks
 *	@top:	  place to the (detached) top of branch
 *
 *	This is a helper function used by ext4_truncate().
 *
 *	When we do truncate() we may have to clean the ends of several
 *	indirect blocks but leave the blocks themselves alive. Block is
 *	partially truncated if some data below the new i_size is refered
 *	from it (and it is on the path to the first completely truncated
 *	data block, indeed).  We have to free the top of that path along
 *	with everything to the right of the path. Since no allocation
 *	past the truncation point is possible until ext4_truncate()
 *	finishes, we may safely do the latter, but top of branch may
 *	require special attention - pageout below the truncation point
 *	might try to populate it.
 *
 *	We atomically detach the top of branch from the tree, store the
 *	block number of its root in *@top, pointers to buffer_heads of
 *	partially truncated blocks - in @chain[].bh and pointers to
 *	their last elements that should not be removed - in
 *	@chain[].p. Return value is the pointer to last filled element
 *	of @chain.
 *
 *	The work left to caller to do the actual freeing of subtrees:
 *		a) free the subtree starting from *@top
 *		b) free the subtrees whose roots are stored in
 *			(@chain[i].p+1 .. end of @chain[i].bh->b_data)
 *		c) free the subtrees growing from the inode past the @chain[0].
 *			(no partially truncated stuff there).  */

static Indirect *ext4_find_shared(struct inode *inode, int depth,
			int offsets[4], Indirect chain[4], __le32 *top)
{
	Indirect *partial, *p;
	int k, err;

	*top = 0;
	/* Make k index the deepest non-null offest + 1 */
	for (k = depth; k > 1 && !offsets[k-1]; k--)
		;
	partial = ext4_get_branch(inode, k, offsets, chain, &err);
	/* Writer: pointers */
	if (!partial)
		partial = chain + k-1;
	/*
	 * If the branch acquired continuation since we've looked at it -
	 * fine, it should all survive and (new) top doesn't belong to us.
	 */
	if (!partial->key && *partial->p)
		/* Writer: end */
		goto no_top;
	for (p=partial; p>chain && all_zeroes((__le32*)p->bh->b_data,p->p); p--)
		;
	/*
	 * OK, we've found the last block that must survive. The rest of our
	 * branch should be detached before unlocking. However, if that rest
	 * of branch is all ours and does not grow immediately from the inode
	 * it's easier to cheat and just decrement partial->p.
	 */
	if (p == chain + k - 1 && p > chain) {
		p->p--;
	} else {
		*top = *p->p;
		/* Nope, don't do this in ext4.  Must leave the tree intact */
#if 0
		*p->p = 0;
#endif
	}
	/* Writer: end */

	while(partial > p) {
		brelse(partial->bh);
		partial--;
	}
no_top:
	return partial;
}

/*
 * Zero a number of block pointers in either an inode or an indirect block.
 * If we restart the transaction we must again get write access to the
 * indirect block for further modification.
 *
 * We release `count' blocks on disk, but (last - first) may be greater
 * than `count' because there can be holes in there.
 */
static void ext4_clear_blocks(handle_t *handle, struct inode *inode,
		struct buffer_head *bh, ext4_fsblk_t block_to_free,
		unsigned long count, __le32 *first, __le32 *last)
{
	__le32 *p;
	if (try_to_extend_transaction(handle, inode)) {
		if (bh) {
			BUFFER_TRACE(bh, "call ext4_journal_dirty_metadata");
			ext4_journal_dirty_metadata(handle, bh);
		}
		ext4_mark_inode_dirty(handle, inode);
		ext4_journal_test_restart(handle, inode);
		if (bh) {
			BUFFER_TRACE(bh, "retaking write access");
			ext4_journal_get_write_access(handle, bh);
		}
	}

	/*
	 * Any buffers which are on the journal will be in memory. We find
	 * them on the hash table so jbd2_journal_revoke() will run jbd2_journal_forget()
	 * on them.  We've already detached each block from the file, so
	 * bforget() in jbd2_journal_forget() should be safe.
	 *
	 * AKPM: turn on bforget in jbd2_journal_forget()!!!
	 */
	for (p = first; p < last; p++) {
		u32 nr = le32_to_cpu(*p);
		if (nr) {
			struct buffer_head *bh;

			*p = 0;
			bh = sb_find_get_block(inode->i_sb, nr);
			ext4_forget(handle, 0, inode, bh, nr);
		}
	}

	ext4_free_blocks(handle, inode, block_to_free, count);
}

/**
 * ext4_free_data - free a list of data blocks
 * @handle:	handle for this transaction
 * @inode:	inode we are dealing with
 * @this_bh:	indirect buffer_head which contains *@first and *@last
 * @first:	array of block numbers
 * @last:	points immediately past the end of array
 *
 * We are freeing all blocks refered from that array (numbers are stored as
 * little-endian 32-bit) and updating @inode->i_blocks appropriately.
 *
 * We accumulate contiguous runs of blocks to free.  Conveniently, if these
 * blocks are contiguous then releasing them at one time will only affect one
 * or two bitmap blocks (+ group descriptor(s) and superblock) and we won't
 * actually use a lot of journal space.
 *
 * @this_bh will be %NULL if @first and @last point into the inode's direct
 * block pointers.
 */
static void ext4_free_data(handle_t *handle, struct inode *inode,
			   struct buffer_head *this_bh,
			   __le32 *first, __le32 *last)
{
	ext4_fsblk_t block_to_free = 0;    /* Starting block # of a run */
	unsigned long count = 0;	    /* Number of blocks in the run */
	__le32 *block_to_free_p = NULL;	    /* Pointer into inode/ind
					       corresponding to
					       block_to_free */
	ext4_fsblk_t nr;		    /* Current block # */
	__le32 *p;			    /* Pointer into inode/ind
					       for current block */
	int err;

	if (this_bh) {				/* For indirect block */
		BUFFER_TRACE(this_bh, "get_write_access");
		err = ext4_journal_get_write_access(handle, this_bh);
		/* Important: if we can't update the indirect pointers
		 * to the blocks, we can't free them. */
		if (err)
			return;
	}

	for (p = first; p < last; p++) {
		nr = le32_to_cpu(*p);
		if (nr) {
			/* accumulate blocks to free if they're contiguous */
			if (count == 0) {
				block_to_free = nr;
				block_to_free_p = p;
				count = 1;
			} else if (nr == block_to_free + count) {
				count++;
			} else {
				ext4_clear_blocks(handle, inode, this_bh,
						  block_to_free,
						  count, block_to_free_p, p);
				block_to_free = nr;
				block_to_free_p = p;
				count = 1;
			}
		}
	}

	if (count > 0)
		ext4_clear_blocks(handle, inode, this_bh, block_to_free,
				  count, block_to_free_p, p);

	if (this_bh) {
		BUFFER_TRACE(this_bh, "call ext4_journal_dirty_metadata");
		ext4_journal_dirty_metadata(handle, this_bh);
	}
}

/**
 *	ext4_free_branches - free an array of branches
 *	@handle: JBD handle for this transaction
 *	@inode:	inode we are dealing with
 *	@parent_bh: the buffer_head which contains *@first and *@last
 *	@first:	array of block numbers
 *	@last:	pointer immediately past the end of array
 *	@depth:	depth of the branches to free
 *
 *	We are freeing all blocks refered from these branches (numbers are
 *	stored as little-endian 32-bit) and updating @inode->i_blocks
 *	appropriately.
 */
static void ext4_free_branches(handle_t *handle, struct inode *inode,
			       struct buffer_head *parent_bh,
			       __le32 *first, __le32 *last, int depth)
{
	ext4_fsblk_t nr;
	__le32 *p;

	if (is_handle_aborted(handle))
		return;

	if (depth--) {
		struct buffer_head *bh;
		int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb);
		p = last;
		while (--p >= first) {
			nr = le32_to_cpu(*p);
			if (!nr)
				continue;		/* A hole */

			/* Go read the buffer for the next level down */
			bh = sb_bread(inode->i_sb, nr);

			/*
			 * A read failure? Report error and clear slot
			 * (should be rare).
			 */
			if (!bh) {
				ext4_error(inode->i_sb, "ext4_free_branches",
					   "Read failure, inode=%lu, block=%llu",
					   inode->i_ino, nr);
				continue;
			}

			/* This zaps the entire block.  Bottom up. */
			BUFFER_TRACE(bh, "free child branches");
			ext4_free_branches(handle, inode, bh,
					   (__le32*)bh->b_data,
					   (__le32*)bh->b_data + addr_per_block,
					   depth);

			/*
			 * We've probably journalled the indirect block several
			 * times during the truncate.  But it's no longer
			 * needed and we now drop it from the transaction via
			 * jbd2_journal_revoke().
			 *
			 * That's easy if it's exclusively part of this
			 * transaction.  But if it's part of the committing
			 * transaction then jbd2_journal_forget() will simply
			 * brelse() it.  That means that if the underlying
			 * block is reallocated in ext4_get_block(),
			 * unmap_underlying_metadata() will find this block
			 * and will try to get rid of it.  damn, damn.
			 *
			 * If this block has already been committed to the
			 * journal, a revoke record will be written.  And
			 * revoke records must be emitted *before* clearing
			 * this block's bit in the bitmaps.
			 */
			ext4_forget(handle, 1, inode, bh, bh->b_blocknr);

			/*
			 * Everything below this this pointer has been
			 * released.  Now let this top-of-subtree go.
			 *
			 * We want the freeing of this indirect block to be
			 * atomic in the journal with the updating of the
			 * bitmap block which owns it.  So make some room in
			 * the journal.
			 *
			 * We zero the parent pointer *after* freeing its
			 * pointee in the bitmaps, so if extend_transaction()
			 * for some reason fails to put the bitmap changes and
			 * the release into the same transaction, recovery
			 * will merely complain about releasing a free block,
			 * rather than leaking blocks.
			 */
			if (is_handle_ab