[people/ms/linux.git] / fs / f2fs / data.c

/*
 * fs/f2fs/data.c
 *
 * Copyright (c) 2012 Samsung Electronics Co., Ltd.
 *             http://www.samsung.com/
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */
#include <linux/fs.h>
#include <linux/f2fs_fs.h>
#include <linux/buffer_head.h>
#include <linux/mpage.h>
#include <linux/writeback.h>
#include <linux/backing-dev.h>
#include <linux/blkdev.h>
#include <linux/bio.h>
#include <linux/prefetch.h>

#include "f2fs.h"
#include "node.h"
#include "segment.h"
#include <trace/events/f2fs.h>

/*
 * Lock ordering for the change of data block address:
 * ->data_page
 *  ->node_page
 *    update block addresses in the node page
 */
static void __set_data_blkaddr(struct dnode_of_data *dn, block_t new_addr)
{
	struct f2fs_node *rn;
	__le32 *addr_array;
	struct page *node_page = dn->node_page;
	unsigned int ofs_in_node = dn->ofs_in_node;

	wait_on_page_writeback(node_page);

	rn = (struct f2fs_node *)page_address(node_page);

	/* Get physical address of data block */
	addr_array = blkaddr_in_node(rn);
	addr_array[ofs_in_node] = cpu_to_le32(new_addr);
	set_page_dirty(node_page);
}

int reserve_new_block(struct dnode_of_data *dn)
{
	struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);

	if (is_inode_flag_set(F2FS_I(dn->inode), FI_NO_ALLOC))
		return -EPERM;
	if (!inc_valid_block_count(sbi, dn->inode, 1))
		return -ENOSPC;

	trace_f2fs_reserve_new_block(dn->inode, dn->nid, dn->ofs_in_node);

	__set_data_blkaddr(dn, NEW_ADDR);
	dn->data_blkaddr = NEW_ADDR;
	sync_inode_page(dn);
	return 0;
}

static int check_extent_cache(struct inode *inode, pgoff_t pgofs,
					struct buffer_head *bh_result)
{
	struct f2fs_inode_info *fi = F2FS_I(inode);
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	pgoff_t start_fofs, end_fofs;
	block_t start_blkaddr;

	read_lock(&fi->ext.ext_lock);
	if (fi->ext.len == 0) {
		read_unlock(&fi->ext.ext_lock);
		return 0;
	}

	sbi->total_hit_ext++;
	start_fofs = fi->ext.fofs;
	end_fofs = fi->ext.fofs + fi->ext.len - 1;
	start_blkaddr = fi->ext.blk_addr;

	if (pgofs >= start_fofs && pgofs <= end_fofs) {
		unsigned int blkbits = inode->i_sb->s_blocksize_bits;
		size_t count;

		clear_buffer_new(bh_result);
		map_bh(bh_result, inode->i_sb,
				start_blkaddr + pgofs - start_fofs);
		count = end_fofs - pgofs + 1;
		if (count < (UINT_MAX >> blkbits))
			bh_result->b_size = (count << blkbits);
		else
			bh_result->b_size = UINT_MAX;

		sbi->read_hit_ext++;
		read_unlock(&fi->ext.ext_lock);
		return 1;
	}
	read_unlock(&fi->ext.ext_lock);
	return 0;
}

void update_extent_cache(block_t blk_addr, struct dnode_of_data *dn)
{
	struct f2fs_inode_info *fi = F2FS_I(dn->inode);
	pgoff_t fofs, start_fofs, end_fofs;
	block_t start_blkaddr, end_blkaddr;

	BUG_ON(blk_addr == NEW_ADDR);
	fofs = start_bidx_of_node(ofs_of_node(dn->node_page)) + dn->ofs_in_node;

	/* Update the page address in the parent node */
	__set_data_blkaddr(dn, blk_addr);

	write_lock(&fi->ext.ext_lock);

	start_fofs = fi->ext.fofs;
	end_fofs = fi->ext.fofs + fi->ext.len - 1;
	start_blkaddr = fi->ext.blk_addr;
	end_blkaddr = fi->ext.blk_addr + fi->ext.len - 1;

	/* Drop and initialize the matched extent */
	if (fi->ext.len == 1 && fofs == start_fofs)
		fi->ext.len = 0;

	/* Initial extent */
	if (fi->ext.len == 0) {
		if (blk_addr != NULL_ADDR) {
			fi->ext.fofs = fofs;
			fi->ext.blk_addr = blk_addr;
			fi->ext.len = 1;
		}
		goto end_update;
	}

	/* Front merge */
	if (fofs == start_fofs - 1 && blk_addr == start_blkaddr - 1) {
		fi->ext.fofs--;
		fi->ext.blk_addr--;
		fi->ext.len++;
		goto end_update;
	}

	/* Back merge */
	if (fofs == end_fofs + 1 && blk_addr == end_blkaddr + 1) {
		fi->ext.len++;
		goto end_update;
	}

	/* Split the existing extent */
	if (fi->ext.len > 1 &&
		fofs >= start_fofs && fofs <= end_fofs) {
		if ((end_fofs - fofs) < (fi->ext.len >> 1)) {
			fi->ext.len = fofs - start_fofs;
		} else {
			fi->ext.fofs = fofs + 1;
			fi->ext.blk_addr = start_blkaddr +
					fofs - start_fofs + 1;
			fi->ext.len -= fofs - start_fofs + 1;
		}
		goto end_update;
	}
	write_unlock(&fi->ext.ext_lock);
	return;

end_update:
	write_unlock(&fi->ext.ext_lock);
	sync_inode_page(dn);
	return;
}

struct page *find_data_page(struct inode *inode, pgoff_t index, bool sync)
{
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	struct address_space *mapping = inode->i_mapping;
	struct dnode_of_data dn;
	struct page *page;
	int err;

	page = find_get_page(mapping, index);
	if (page && PageUptodate(page))
		return page;
	f2fs_put_page(page, 0);

	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, index, LOOKUP_NODE);
	if (err)
		return ERR_PTR(err);
	f2fs_put_dnode(&dn);

	if (dn.data_blkaddr == NULL_ADDR)
		return ERR_PTR(-ENOENT);

	/* By fallocate(), there is no cached page, but with NEW_ADDR */
	if (dn.data_blkaddr == NEW_ADDR)
		return ERR_PTR(-EINVAL);

	page = grab_cache_page(mapping, index);
	if (!page)
		return ERR_PTR(-ENOMEM);

	if (PageUptodate(page)) {
		unlock_page(page);
		return page;
	}

	err = f2fs_readpage(sbi, page, dn.data_blkaddr,
					sync ? READ_SYNC : READA);
	if (sync) {
		wait_on_page_locked(page);
		if (!PageUptodate(page)) {
			f2fs_put_page(page, 0);
			return ERR_PTR(-EIO);
		}
	}
	return page;
}

/*
 * If it tries to access a hole, return an error.
 * Because, the callers, functions in dir.c and GC, should be able to know
 * whether this page exists or not.
 */
struct page *get_lock_data_page(struct inode *inode, pgoff_t index)
{
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	struct address_space *mapping = inode->i_mapping;
	struct dnode_of_data dn;
	struct page *page;
	int err;

	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, index, LOOKUP_NODE);
	if (err)
		return ERR_PTR(err);
	f2fs_put_dnode(&dn);

	if (dn.data_blkaddr == NULL_ADDR)
		return ERR_PTR(-ENOENT);

	page = grab_cache_page(mapping, index);
	if (!page)
		return ERR_PTR(-ENOMEM);

	if (PageUptodate(page))
		return page;

	BUG_ON(dn.data_blkaddr == NEW_ADDR);
	BUG_ON(dn.data_blkaddr == NULL_ADDR);

	err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
	if (err)
		return ERR_PTR(err);

	lock_page(page);
	if (!PageUptodate(page)) {
		f2fs_put_page(page, 1);
		return ERR_PTR(-EIO);
	}
	return page;
}

/*
 * Caller ensures that this data page is never allocated.
 * A new zero-filled data page is allocated in the page cache.
 *
 * Also, caller should grab and release a mutex by calling mutex_lock_op() and
 * mutex_unlock_op().
 */
struct page *get_new_data_page(struct inode *inode, pgoff_t index,
						bool new_i_size)
{
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	struct address_space *mapping = inode->i_mapping;
	struct page *page;
	struct dnode_of_data dn;
	int err;

	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, index, ALLOC_NODE);
	if (err)
		return ERR_PTR(err);

	if (dn.data_blkaddr == NULL_ADDR) {
		if (reserve_new_block(&dn)) {
			f2fs_put_dnode(&dn);
			return ERR_PTR(-ENOSPC);
		}
	}
	f2fs_put_dnode(&dn);

	page = grab_cache_page(mapping, index);
	if (!page)
		return ERR_PTR(-ENOMEM);

	if (PageUptodate(page))
		return page;

	if (dn.data_blkaddr == NEW_ADDR) {
		zero_user_segment(page, 0, PAGE_CACHE_SIZE);
		SetPageUptodate(page);
	} else {
		err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
		if (err)
			return ERR_PTR(err);
		lock_page(page);
		if (!PageUptodate(page)) {
			f2fs_put_page(page, 1);
			return ERR_PTR(-EIO);
		}
	}

	if (new_i_size &&
		i_size_read(inode) < ((index + 1) << PAGE_CACHE_SHIFT)) {
		i_size_write(inode, ((index + 1) << PAGE_CACHE_SHIFT));
		mark_inode_dirty_sync(inode);
	}
	return page;
}

static void read_end_io(struct bio *bio, int err)
{
	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1;

	do {
		struct page *page = bvec->bv_page;

		if (--bvec >= bio->bi_io_vec)
			prefetchw(&bvec->bv_page->flags);

		if (uptodate) {
			SetPageUptodate(page);
		} else {
			ClearPageUptodate(page);
			SetPageError(page);
		}
		unlock_page(page);
	} while (bvec >= bio->bi_io_vec);
	kfree(bio->bi_private);
	bio_put(bio);
}

/*
 * Fill the locked page with data located in the block address.
 * Return unlocked page.
 */
int f2fs_readpage(struct f2fs_sb_info *sbi, struct page *page,
					block_t blk_addr, int type)
{
	struct block_device *bdev = sbi->sb->s_bdev;
	struct bio *bio;

	trace_f2fs_readpage(page, blk_addr, type);

	down_read(&sbi->bio_sem);

	/* Allocate a new bio */
	bio = f2fs_bio_alloc(bdev, 1);

	/* Initialize the bio */
	bio->bi_sector = SECTOR_FROM_BLOCK(sbi, blk_addr);
	bio->bi_end_io = read_end_io;

	if (bio_add_page(bio, page, PAGE_CACHE_SIZE, 0) < PAGE_CACHE_SIZE) {
		kfree(bio->bi_private);
		bio_put(bio);
		up_read(&sbi->bio_sem);
		f2fs_put_page(page, 1);
		return -EFAULT;
	}

	submit_bio(type, bio);
	up_read(&sbi->bio_sem);
	return 0;
}

/*
 * This function should be used by the data read flow only where it
 * does not check the "create" flag that indicates block allocation.
 * The reason for this special functionality is to exploit VFS readahead
 * mechanism.
 */
static int get_data_block_ro(struct inode *inode, sector_t iblock,
			struct buffer_head *bh_result, int create)
{
	unsigned int blkbits = inode->i_sb->s_blocksize_bits;
	unsigned maxblocks = bh_result->b_size >> blkbits;
	struct dnode_of_data dn;
	pgoff_t pgofs;
	int err;

	/* Get the page offset from the block offset(iblock) */
	pgofs =	(pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits));

	if (check_extent_cache(inode, pgofs, bh_result)) {
		trace_f2fs_get_data_block(inode, iblock, bh_result, 0);
		return 0;
	}

	/* When reading holes, we need its node page */
	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
	if (err) {
		trace_f2fs_get_data_block(inode, iblock, bh_result, err);
		return (err == -ENOENT) ? 0 : err;
	}

	/* It does not support data allocation */
	BUG_ON(create);

	if (dn.data_blkaddr != NEW_ADDR && dn.data_blkaddr != NULL_ADDR) {
		int i;
		unsigned int end_offset;

		end_offset = IS_INODE(dn.node_page) ?
				ADDRS_PER_INODE :
				ADDRS_PER_BLOCK;

		clear_buffer_new(bh_result);

		/* Give more consecutive addresses for the read ahead */
		for (i = 0; i < end_offset - dn.ofs_in_node; i++)
			if (((datablock_addr(dn.node_page,
							dn.ofs_in_node + i))
				!= (dn.data_blkaddr + i)) || maxblocks == i)
				break;
		map_bh(bh_result, inode->i_sb, dn.data_blkaddr);
		bh_result->b_size = (i << blkbits);
	}
	f2fs_put_dnode(&dn);
	trace_f2fs_get_data_block(inode, iblock, bh_result, 0);
	return 0;
}

static int f2fs_read_data_page(struct file *file, struct page *page)
{
	return mpage_readpage(page, get_data_block_ro);
}

static int f2fs_read_data_pages(struct file *file,
			struct address_space *mapping,
			struct list_head *pages, unsigned nr_pages)
{
	return mpage_readpages(mapping, pages, nr_pages, get_data_block_ro);
}

int do_write_data_page(struct page *page)
{
	struct inode *inode = page->mapping->host;
	block_t old_blk_addr, new_blk_addr;
	struct dnode_of_data dn;
	int err = 0;

	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, page->index, LOOKUP_NODE);
	if (err)
		return err;

	old_blk_addr = dn.data_blkaddr;

	/* This page is already truncated */
	if (old_blk_addr == NULL_ADDR)
		goto out_writepage;

	set_page_writeback(page);

	/*
	 * If current allocation needs SSR,
	 * it had better in-place writes for updated data.
	 */
	if (old_blk_addr != NEW_ADDR && !is_cold_data(page) &&
				need_inplace_update(inode)) {
		rewrite_data_page(F2FS_SB(inode->i_sb), page,
						old_blk_addr);
	} else {
		write_data_page(inode, page, &dn,
				old_blk_addr, &new_blk_addr);
		update_extent_cache(new_blk_addr, &dn);
	}
out_writepage:
	f2fs_put_dnode(&dn);
	return err;
}

static int f2fs_write_data_page(struct page *page,
					struct writeback_control *wbc)
{
	struct inode *inode = page->mapping->host;
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	loff_t i_size = i_size_read(inode);
	const pgoff_t end_index = ((unsigned long long) i_size)
							>> PAGE_CACHE_SHIFT;
	unsigned offset;
	bool need_balance_fs = false;
	int err = 0;

	if (page->index < end_index)
		goto write;

	/*
	 * If the offset is out-of-range of file size,
	 * this page does not have to be written to disk.
	 */
	offset = i_size & (PAGE_CACHE_SIZE - 1);
	if ((page->index >= end_index + 1) || !offset) {
		if (S_ISDIR(inode->i_mode)) {
			dec_page_count(sbi, F2FS_DIRTY_DENTS);
			inode_dec_dirty_dents(inode);
		}
		goto out;
	}

	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
write:
	if (sbi->por_doing) {
		err = AOP_WRITEPAGE_ACTIVATE;
		goto redirty_out;
	}

	/* Dentry blocks are controlled by checkpoint */
	if (S_ISDIR(inode->i_mode)) {
		dec_page_count(sbi, F2FS_DIRTY_DENTS);
		inode_dec_dirty_dents(inode);
		err = do_write_data_page(page);
	} else {
		int ilock = mutex_lock_op(sbi);
		err = do_write_data_page(page);
		mutex_unlock_op(sbi, ilock);
		need_balance_fs = true;
	}
	if (err == -ENOENT)
		goto out;
	else if (err)
		goto redirty_out;

	if (wbc->for_reclaim)
		f2fs_submit_bio(sbi, DATA, true);

	clear_cold_data(page);
out:
	unlock_page(page);
	if (need_balance_fs)
		f2fs_balance_fs(sbi);
	return 0;

redirty_out:
	wbc->pages_skipped++;
	set_page_dirty(page);
	return err;
}

#define MAX_DESIRED_PAGES_WP	4096

static int __f2fs_writepage(struct page *page, struct writeback_control *wbc,
			void *data)
{
	struct address_space *mapping = data;
	int ret = mapping->a_ops->writepage(page, wbc);
	mapping_set_error(mapping, ret);
	return ret;
}

static int f2fs_write_data_pages(struct address_space *mapping,
			    struct writeback_control *wbc)
{
	struct inode *inode = mapping->host;
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	int ret;
	long excess_nrtw = 0, desired_nrtw;

	/* deal with chardevs and other special file */
	if (!mapping->a_ops->writepage)
		return 0;

	if (wbc->nr_to_write < MAX_DESIRED_PAGES_WP) {
		desired_nrtw = MAX_DESIRED_PAGES_WP;
		excess_nrtw = desired_nrtw - wbc->nr_to_write;
		wbc->nr_to_write = desired_nrtw;
	}

	if (!S_ISDIR(inode->i_mode))
		mutex_lock(&sbi->writepages);
	ret = write_cache_pages(mapping, wbc, __f2fs_writepage, mapping);
	if (!S_ISDIR(inode->i_mode))
		mutex_unlock(&sbi->writepages);
	f2fs_submit_bio(sbi, DATA, (wbc->sync_mode == WB_SYNC_ALL));

	remove_dirty_dir_inode(inode);

	wbc->nr_to_write -= excess_nrtw;
	return ret;
}

static int f2fs_write_begin(struct file *file, struct address_space *mapping,
		loff_t pos, unsigned len, unsigned flags,
		struct page **pagep, void **fsdata)
{
	struct inode *inode = mapping->host;
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	struct page *page;
	pgoff_t index = ((unsigned long long) pos) >> PAGE_CACHE_SHIFT;
	struct dnode_of_data dn;
	int err = 0;
	int ilock;

	/* for nobh_write_end */
	*fsdata = NULL;

	f2fs_balance_fs(sbi);

	page = grab_cache_page_write_begin(mapping, index, flags);
	if (!page)
		return -ENOMEM;
	*pagep = page;

	ilock = mutex_lock_op(sbi);

	set_new_dnode(&dn, inode, NULL, NULL, 0);
	err = get_dnode_of_data(&dn, index, ALLOC_NODE);
	if (err)
		goto err;

	if (dn.data_blkaddr == NULL_ADDR)
		err = reserve_new_block(&dn);

	f2fs_put_dnode(&dn);
	if (err)
		goto err;

	mutex_unlock_op(sbi, ilock);

	if ((len == PAGE_CACHE_SIZE) || PageUptodate(page))
		return 0;

	if ((pos & PAGE_CACHE_MASK) >= i_size_read(inode)) {
		unsigned start = pos & (PAGE_CACHE_SIZE - 1);
		unsigned end = start + len;

		/* Reading beyond i_size is simple: memset to zero */
		zero_user_segments(page, 0, start, end, PAGE_CACHE_SIZE);
		goto out;
	}

	if (dn.data_blkaddr == NEW_ADDR) {
		zero_user_segment(page, 0, PAGE_CACHE_SIZE);
	} else {
		err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
		if (err)
			return err;
		lock_page(page);
		if (!PageUptodate(page)) {
			f2fs_put_page(page, 1);
			return -EIO;
		}
	}
out:
	SetPageUptodate(page);
	clear_cold_data(page);
	return 0;

err:
	mutex_unlock_op(sbi, ilock);
	f2fs_put_page(page, 1);
	return err;
}

static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb,
		const struct iovec *iov, loff_t offset, unsigned long nr_segs)
{
	struct file *file = iocb->ki_filp;
	struct inode *inode = file->f_mapping->host;

	if (rw == WRITE)
		return 0;

	/* Needs synchronization with the cleaner */
	return blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
						  get_data_block_ro);
}

static void f2fs_invalidate_data_page(struct page *page, unsigned long offset)
{
	struct inode *inode = page->mapping->host;
	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	if (S_ISDIR(inode->i_mode) && PageDirty(page)) {
		dec_page_count(sbi, F2FS_DIRTY_DENTS);
		inode_dec_dirty_dents(inode);
	}
	ClearPagePrivate(page);
}

static int f2fs_release_data_page(struct page *page, gfp_t wait)
{
	ClearPagePrivate(page);
	return 1;
}

static int f2fs_set_data_page_dirty(struct page *page)
{
	struct address_space *mapping = page->mapping;
	struct inode *inode = mapping->host;

	SetPageUptodate(page);
	if (!PageDirty(page)) {
		__set_page_dirty_nobuffers(page);
		set_dirty_dir_page(inode, page);
		return 1;
	}
	return 0;
}

static sector_t f2fs_bmap(struct address_space *mapping, sector_t block)
{
	return generic_block_bmap(mapping, block, get_data_block_ro);
}

const struct address_space_operations f2fs_dblock_aops = {
	.readpage	= f2fs_read_data_page,
	.readpages	= f2fs_read_data_pages,
	.writepage	= f2fs_write_data_page,
	.writepages	= f2fs_write_data_pages,
	.write_begin	= f2fs_write_begin,
	.write_end	= nobh_write_end,
	.set_page_dirty	= f2fs_set_data_page_dirty,
	.invalidatepage	= f2fs_invalidate_data_page,
	.releasepage	= f2fs_release_data_page,
	.direct_IO	= f2fs_direct_IO,
	.bmap		= f2fs_bmap,
};
Commit	Line	Data
0a8165d7	1	/*
eb47b800 JK	2	* fs/f2fs/data.c
	3	*
	4	* Copyright (c) 2012 Samsung Electronics Co., Ltd.
	5	* http://www.samsung.com/
	6	*
	7	* This program is free software; you can redistribute it and/or modify
	8	* it under the terms of the GNU General Public License version 2 as
	9	* published by the Free Software Foundation.
	10	*/
	11	#include <linux/fs.h>
	12	#include <linux/f2fs_fs.h>
	13	#include <linux/buffer_head.h>
	14	#include <linux/mpage.h>
	15	#include <linux/writeback.h>
	16	#include <linux/backing-dev.h>
	17	#include <linux/blkdev.h>
	18	#include <linux/bio.h>
690e4a3e	19	#include <linux/prefetch.h>
eb47b800 JK	20
	21	#include "f2fs.h"
	22	#include "node.h"
	23	#include "segment.h"
848753aa	24	#include <trace/events/f2fs.h>
eb47b800	25
0a8165d7	26	/*
eb47b800 JK	27	* Lock ordering for the change of data block address:
	28	* ->data_page
	29	* ->node_page
	30	* update block addresses in the node page
	31	*/
	32	static void __set_data_blkaddr(struct dnode_of_data *dn, block_t new_addr)
	33	{
	34	struct f2fs_node *rn;
	35	__le32 *addr_array;
	36	struct page *node_page = dn->node_page;
	37	unsigned int ofs_in_node = dn->ofs_in_node;
	38
	39	wait_on_page_writeback(node_page);
	40
	41	rn = (struct f2fs_node *)page_address(node_page);
	42
	43	/* Get physical address of data block */
	44	addr_array = blkaddr_in_node(rn);
	45	addr_array[ofs_in_node] = cpu_to_le32(new_addr);
	46	set_page_dirty(node_page);
	47	}
	48
	49	int reserve_new_block(struct dnode_of_data *dn)
	50	{
	51	struct f2fs_sb_info *sbi = F2FS_SB(dn->inode->i_sb);
	52
	53	if (is_inode_flag_set(F2FS_I(dn->inode), FI_NO_ALLOC))
	54	return -EPERM;
	55	if (!inc_valid_block_count(sbi, dn->inode, 1))
	56	return -ENOSPC;
	57
c01e2853 NJ	58	trace_f2fs_reserve_new_block(dn->inode, dn->nid, dn->ofs_in_node);
c01e2853 NJ	59
eb47b800 JK	60	__set_data_blkaddr(dn, NEW_ADDR);
	61	dn->data_blkaddr = NEW_ADDR;
	62	sync_inode_page(dn);
	63	return 0;
	64	}
	65
	66	static int check_extent_cache(struct inode *inode, pgoff_t pgofs,
	67	struct buffer_head *bh_result)
	68	{
	69	struct f2fs_inode_info *fi = F2FS_I(inode);
	70	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	71	pgoff_t start_fofs, end_fofs;
	72	block_t start_blkaddr;
	73
	74	read_lock(&fi->ext.ext_lock);
	75	if (fi->ext.len == 0) {
	76	read_unlock(&fi->ext.ext_lock);
	77	return 0;
	78	}
	79
	80	sbi->total_hit_ext++;
	81	start_fofs = fi->ext.fofs;
	82	end_fofs = fi->ext.fofs + fi->ext.len - 1;
	83	start_blkaddr = fi->ext.blk_addr;
	84
	85	if (pgofs >= start_fofs && pgofs <= end_fofs) {
	86	unsigned int blkbits = inode->i_sb->s_blocksize_bits;
	87	size_t count;
	88
	89	clear_buffer_new(bh_result);
	90	map_bh(bh_result, inode->i_sb,
	91	start_blkaddr + pgofs - start_fofs);
	92	count = end_fofs - pgofs + 1;
	93	if (count < (UINT_MAX >> blkbits))
	94	bh_result->b_size = (count << blkbits);
	95	else
	96	bh_result->b_size = UINT_MAX;
	97
	98	sbi->read_hit_ext++;
	99	read_unlock(&fi->ext.ext_lock);
	100	return 1;
	101	}
	102	read_unlock(&fi->ext.ext_lock);
	103	return 0;
	104	}
	105
	106	void update_extent_cache(block_t blk_addr, struct dnode_of_data *dn)
	107	{
	108	struct f2fs_inode_info *fi = F2FS_I(dn->inode);
	109	pgoff_t fofs, start_fofs, end_fofs;
	110	block_t start_blkaddr, end_blkaddr;
	111
	112	BUG_ON(blk_addr == NEW_ADDR);
	113	fofs = start_bidx_of_node(ofs_of_node(dn->node_page)) + dn->ofs_in_node;
	114
	115	/* Update the page address in the parent node */
	116	__set_data_blkaddr(dn, blk_addr);
	117
	118	write_lock(&fi->ext.ext_lock);
	119
	120	start_fofs = fi->ext.fofs;
	121	end_fofs = fi->ext.fofs + fi->ext.len - 1;
	122	start_blkaddr = fi->ext.blk_addr;
	123	end_blkaddr = fi->ext.blk_addr + fi->ext.len - 1;
124
125	/* Drop and initialize the matched extent */
126	if (fi->ext.len == 1 && fofs == start_fofs)
127	fi->ext.len = 0;
128
129	/* Initial extent */
130	if (fi->ext.len == 0) {
131	if (blk_addr != NULL_ADDR) {
132	fi->ext.fofs = fofs;
133	fi->ext.blk_addr = blk_addr;
134	fi->ext.len = 1;
135	}
136	goto end_update;
137	}
138
6224da87	139	/* Front merge */
eb47b800 JK	140	if (fofs == start_fofs - 1 && blk_addr == start_blkaddr - 1) {
	141	fi->ext.fofs--;
	142	fi->ext.blk_addr--;
	143	fi->ext.len++;
	144	goto end_update;
	145	}
	146
	147	/* Back merge */
	148	if (fofs == end_fofs + 1 && blk_addr == end_blkaddr + 1) {
	149	fi->ext.len++;
	150	goto end_update;
	151	}
	152
	153	/* Split the existing extent */
	154	if (fi->ext.len > 1 &&
	155	fofs >= start_fofs && fofs <= end_fofs) {
	156	if ((end_fofs - fofs) < (fi->ext.len >> 1)) {
	157	fi->ext.len = fofs - start_fofs;
	158	} else {
	159	fi->ext.fofs = fofs + 1;
	160	fi->ext.blk_addr = start_blkaddr +
	161	fofs - start_fofs + 1;
	162	fi->ext.len -= fofs - start_fofs + 1;
	163	}
	164	goto end_update;
	165	}
	166	write_unlock(&fi->ext.ext_lock);
	167	return;
	168
	169	end_update:
	170	write_unlock(&fi->ext.ext_lock);
	171	sync_inode_page(dn);
	172	return;
	173	}
	174
c718379b	175	struct page find_data_page(struct inode inode, pgoff_t index, bool sync)
eb47b800 JK	176	{
	177	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	178	struct address_space *mapping = inode->i_mapping;
	179	struct dnode_of_data dn;
	180	struct page *page;
	181	int err;
	182
	183	page = find_get_page(mapping, index);
	184	if (page && PageUptodate(page))
	185	return page;
	186	f2fs_put_page(page, 0);
	187
	188	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	189	err = get_dnode_of_data(&dn, index, LOOKUP_NODE);
eb47b800 JK	190	if (err)
	191	return ERR_PTR(err);
	192	f2fs_put_dnode(&dn);
	193
	194	if (dn.data_blkaddr == NULL_ADDR)
	195	return ERR_PTR(-ENOENT);
	196
	197	/* By fallocate(), there is no cached page, but with NEW_ADDR */
	198	if (dn.data_blkaddr == NEW_ADDR)
	199	return ERR_PTR(-EINVAL);
	200
	201	page = grab_cache_page(mapping, index);
	202	if (!page)
	203	return ERR_PTR(-ENOMEM);
	204
393ff91f JK	205	if (PageUptodate(page)) {
	206	unlock_page(page);
	207	return page;
	208	}
	209
c718379b JK	210	err = f2fs_readpage(sbi, page, dn.data_blkaddr,
	211	sync ? READ_SYNC : READA);
	212	if (sync) {
	213	wait_on_page_locked(page);
	214	if (!PageUptodate(page)) {
	215	f2fs_put_page(page, 0);
	216	return ERR_PTR(-EIO);
	217	}
eb47b800	218	}
eb47b800 JK	219	return page;
	220	}
	221
0a8165d7	222	/*
eb47b800 JK	223	* If it tries to access a hole, return an error.
	224	* Because, the callers, functions in dir.c and GC, should be able to know
	225	* whether this page exists or not.
	226	*/
	227	struct page get_lock_data_page(struct inode inode, pgoff_t index)
	228	{
	229	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	230	struct address_space *mapping = inode->i_mapping;
	231	struct dnode_of_data dn;
	232	struct page *page;
	233	int err;
	234
	235	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	236	err = get_dnode_of_data(&dn, index, LOOKUP_NODE);
eb47b800 JK	237	if (err)
	238	return ERR_PTR(err);
	239	f2fs_put_dnode(&dn);
	240
	241	if (dn.data_blkaddr == NULL_ADDR)
	242	return ERR_PTR(-ENOENT);
	243
	244	page = grab_cache_page(mapping, index);
	245	if (!page)
	246	return ERR_PTR(-ENOMEM);
	247
	248	if (PageUptodate(page))
	249	return page;
	250
	251	BUG_ON(dn.data_blkaddr == NEW_ADDR);
	252	BUG_ON(dn.data_blkaddr == NULL_ADDR);
	253
	254	err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
393ff91f	255	if (err)
eb47b800	256	return ERR_PTR(err);
393ff91f JK	257
	258	lock_page(page);
	259	if (!PageUptodate(page)) {
	260	f2fs_put_page(page, 1);
	261	return ERR_PTR(-EIO);
eb47b800 JK	262	}
	263	return page;
	264	}
	265
0a8165d7	266	/*
eb47b800 JK	267	* Caller ensures that this data page is never allocated.
eb47b800 JK	268	* A new zero-filled data page is allocated in the page cache.
39936837 JK	269	*
	270	* Also, caller should grab and release a mutex by calling mutex_lock_op() and
	271	* mutex_unlock_op().
eb47b800 JK	272	*/
	273	struct page get_new_data_page(struct inode inode, pgoff_t index,
	274	bool new_i_size)
	275	{
	276	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	277	struct address_space *mapping = inode->i_mapping;
	278	struct page *page;
	279	struct dnode_of_data dn;
	280	int err;
	281
	282	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	283	err = get_dnode_of_data(&dn, index, ALLOC_NODE);
eb47b800 JK	284	if (err)
	285	return ERR_PTR(err);
	286
	287	if (dn.data_blkaddr == NULL_ADDR) {
	288	if (reserve_new_block(&dn)) {
	289	f2fs_put_dnode(&dn);
	290	return ERR_PTR(-ENOSPC);
	291	}
	292	}
	293	f2fs_put_dnode(&dn);
	294
	295	page = grab_cache_page(mapping, index);
	296	if (!page)
	297	return ERR_PTR(-ENOMEM);
	298
	299	if (PageUptodate(page))
	300	return page;
	301
	302	if (dn.data_blkaddr == NEW_ADDR) {
	303	zero_user_segment(page, 0, PAGE_CACHE_SIZE);
393ff91f	304	SetPageUptodate(page);
eb47b800 JK	305	} else {
eb47b800 JK	306	err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
393ff91f	307	if (err)
eb47b800	308	return ERR_PTR(err);
393ff91f JK	309	lock_page(page);
	310	if (!PageUptodate(page)) {
	311	f2fs_put_page(page, 1);
	312	return ERR_PTR(-EIO);
eb47b800 JK	313	}
eb47b800 JK	314	}
eb47b800 JK	315
	316	if (new_i_size &&
	317	i_size_read(inode) < ((index + 1) << PAGE_CACHE_SHIFT)) {
	318	i_size_write(inode, ((index + 1) << PAGE_CACHE_SHIFT));
	319	mark_inode_dirty_sync(inode);
	320	}
	321	return page;
	322	}
	323
	324	static void read_end_io(struct bio *bio, int err)
	325	{
	326	const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	327	struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1;
	328
	329	do {
	330	struct page *page = bvec->bv_page;
	331
	332	if (--bvec >= bio->bi_io_vec)
	333	prefetchw(&bvec->bv_page->flags);
	334
	335	if (uptodate) {
	336	SetPageUptodate(page);
	337	} else {
	338	ClearPageUptodate(page);
	339	SetPageError(page);
	340	}
	341	unlock_page(page);
	342	} while (bvec >= bio->bi_io_vec);
	343	kfree(bio->bi_private);
	344	bio_put(bio);
	345	}
	346
0a8165d7	347	/*
eb47b800	348	* Fill the locked page with data located in the block address.
393ff91f	349	* Return unlocked page.
eb47b800 JK	350	*/
	351	int f2fs_readpage(struct f2fs_sb_info sbi, struct page page,
	352	block_t blk_addr, int type)
	353	{
	354	struct block_device *bdev = sbi->sb->s_bdev;
eb47b800 JK	355	struct bio *bio;
eb47b800 JK	356
848753aa NJ	357	trace_f2fs_readpage(page, blk_addr, type);
848753aa NJ	358
eb47b800 JK	359	down_read(&sbi->bio_sem);
	360
	361	/* Allocate a new bio */
3cd8a239	362	bio = f2fs_bio_alloc(bdev, 1);
eb47b800 JK	363
eb47b800 JK	364	/* Initialize the bio */
3cd8a239	365	bio->bi_sector = SECTOR_FROM_BLOCK(sbi, blk_addr);
eb47b800	366	bio->bi_end_io = read_end_io;
3cd8a239	367
eb47b800 JK	368	if (bio_add_page(bio, page, PAGE_CACHE_SIZE, 0) < PAGE_CACHE_SIZE) {
	369	kfree(bio->bi_private);
	370	bio_put(bio);
	371	up_read(&sbi->bio_sem);
393ff91f	372	f2fs_put_page(page, 1);
eb47b800 JK	373	return -EFAULT;
	374	}
	375
	376	submit_bio(type, bio);
	377	up_read(&sbi->bio_sem);
eb47b800 JK	378	return 0;
	379	}
	380
0a8165d7	381	/*
eb47b800 JK	382	* This function should be used by the data read flow only where it
	383	* does not check the "create" flag that indicates block allocation.
	384	* The reason for this special functionality is to exploit VFS readahead
	385	* mechanism.
	386	*/
	387	static int get_data_block_ro(struct inode *inode, sector_t iblock,
	388	struct buffer_head *bh_result, int create)
	389	{
	390	unsigned int blkbits = inode->i_sb->s_blocksize_bits;
	391	unsigned maxblocks = bh_result->b_size >> blkbits;
	392	struct dnode_of_data dn;
	393	pgoff_t pgofs;
	394	int err;
	395
	396	/* Get the page offset from the block offset(iblock) */
	397	pgofs = (pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits));
	398
848753aa NJ	399	if (check_extent_cache(inode, pgofs, bh_result)) {
848753aa NJ	400	trace_f2fs_get_data_block(inode, iblock, bh_result, 0);
eb47b800	401	return 0;
848753aa	402	}
eb47b800 JK	403
	404	/* When reading holes, we need its node page */
	405	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	406	err = get_dnode_of_data(&dn, pgofs, LOOKUP_NODE_RA);
848753aa NJ	407	if (err) {
848753aa NJ	408	trace_f2fs_get_data_block(inode, iblock, bh_result, err);
eb47b800	409	return (err == -ENOENT) ? 0 : err;
848753aa	410	}
eb47b800 JK	411
	412	/* It does not support data allocation */
	413	BUG_ON(create);
	414
	415	if (dn.data_blkaddr != NEW_ADDR && dn.data_blkaddr != NULL_ADDR) {
	416	int i;
	417	unsigned int end_offset;
	418
	419	end_offset = IS_INODE(dn.node_page) ?
	420	ADDRS_PER_INODE :
	421	ADDRS_PER_BLOCK;
	422
	423	clear_buffer_new(bh_result);
	424
	425	/* Give more consecutive addresses for the read ahead */
	426	for (i = 0; i < end_offset - dn.ofs_in_node; i++)
	427	if (((datablock_addr(dn.node_page,
	428	dn.ofs_in_node + i))
	429	!= (dn.data_blkaddr + i)) \|\| maxblocks == i)
	430	break;
	431	map_bh(bh_result, inode->i_sb, dn.data_blkaddr);
	432	bh_result->b_size = (i << blkbits);
	433	}
	434	f2fs_put_dnode(&dn);
848753aa	435	trace_f2fs_get_data_block(inode, iblock, bh_result, 0);
eb47b800 JK	436	return 0;
	437	}
	438
	439	static int f2fs_read_data_page(struct file file, struct page page)
	440	{
	441	return mpage_readpage(page, get_data_block_ro);
	442	}
	443
	444	static int f2fs_read_data_pages(struct file *file,
	445	struct address_space *mapping,
	446	struct list_head *pages, unsigned nr_pages)
	447	{
	448	return mpage_readpages(mapping, pages, nr_pages, get_data_block_ro);
	449	}
	450
	451	int do_write_data_page(struct page *page)
	452	{
	453	struct inode *inode = page->mapping->host;
eb47b800 JK	454	block_t old_blk_addr, new_blk_addr;
	455	struct dnode_of_data dn;
	456	int err = 0;
	457
	458	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	459	err = get_dnode_of_data(&dn, page->index, LOOKUP_NODE);
eb47b800 JK	460	if (err)
	461	return err;
	462
	463	old_blk_addr = dn.data_blkaddr;
	464
	465	/* This page is already truncated */
	466	if (old_blk_addr == NULL_ADDR)
	467	goto out_writepage;
	468
	469	set_page_writeback(page);
	470
	471	/*
	472	* If current allocation needs SSR,
	473	* it had better in-place writes for updated data.
	474	*/
	475	if (old_blk_addr != NEW_ADDR && !is_cold_data(page) &&
	476	need_inplace_update(inode)) {
	477	rewrite_data_page(F2FS_SB(inode->i_sb), page,
	478	old_blk_addr);
	479	} else {
	480	write_data_page(inode, page, &dn,
	481	old_blk_addr, &new_blk_addr);
	482	update_extent_cache(new_blk_addr, &dn);
eb47b800 JK	483	}
	484	out_writepage:
	485	f2fs_put_dnode(&dn);
	486	return err;
	487	}
	488
	489	static int f2fs_write_data_page(struct page *page,
	490	struct writeback_control *wbc)
	491	{
	492	struct inode *inode = page->mapping->host;
	493	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	494	loff_t i_size = i_size_read(inode);
	495	const pgoff_t end_index = ((unsigned long long) i_size)
	496	>> PAGE_CACHE_SHIFT;
	497	unsigned offset;
39936837	498	bool need_balance_fs = false;
eb47b800 JK	499	int err = 0;
	500
	501	if (page->index < end_index)
39936837	502	goto write;
eb47b800 JK	503
	504	/*
	505	* If the offset is out-of-range of file size,
	506	* this page does not have to be written to disk.
	507	*/
	508	offset = i_size & (PAGE_CACHE_SIZE - 1);
	509	if ((page->index >= end_index + 1) \|\| !offset) {
	510	if (S_ISDIR(inode->i_mode)) {
	511	dec_page_count(sbi, F2FS_DIRTY_DENTS);
	512	inode_dec_dirty_dents(inode);
	513	}
39936837	514	goto out;
eb47b800 JK	515	}
	516
	517	zero_user_segment(page, offset, PAGE_CACHE_SIZE);
39936837 JK	518	write:
	519	if (sbi->por_doing) {
	520	err = AOP_WRITEPAGE_ACTIVATE;
eb47b800	521	goto redirty_out;
39936837	522	}
eb47b800	523
39936837	524	/* Dentry blocks are controlled by checkpoint */
eb47b800 JK	525	if (S_ISDIR(inode->i_mode)) {
	526	dec_page_count(sbi, F2FS_DIRTY_DENTS);
	527	inode_dec_dirty_dents(inode);
39936837 JK	528	err = do_write_data_page(page);
	529	} else {
	530	int ilock = mutex_lock_op(sbi);
	531	err = do_write_data_page(page);
	532	mutex_unlock_op(sbi, ilock);
	533	need_balance_fs = true;
eb47b800	534	}
39936837 JK	535	if (err == -ENOENT)
	536	goto out;
	537	else if (err)
	538	goto redirty_out;
eb47b800 JK	539
	540	if (wbc->for_reclaim)
	541	f2fs_submit_bio(sbi, DATA, true);
	542
eb47b800	543	clear_cold_data(page);
39936837	544	out:
eb47b800	545	unlock_page(page);
39936837	546	if (need_balance_fs)
eb47b800 JK	547	f2fs_balance_fs(sbi);
	548	return 0;
	549
eb47b800 JK	550	redirty_out:
	551	wbc->pages_skipped++;
	552	set_page_dirty(page);
39936837	553	return err;
eb47b800 JK	554	}
	555
	556	#define MAX_DESIRED_PAGES_WP 4096
	557
fa9150a8 NJ	558	static int __f2fs_writepage(struct page page, struct writeback_control wbc,
	559	void *data)
	560	{
	561	struct address_space *mapping = data;
	562	int ret = mapping->a_ops->writepage(page, wbc);
	563	mapping_set_error(mapping, ret);
	564	return ret;
	565	}
	566
25ca923b	567	static int f2fs_write_data_pages(struct address_space *mapping,
eb47b800 JK	568	struct writeback_control *wbc)
	569	{
	570	struct inode *inode = mapping->host;
	571	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	572	int ret;
	573	long excess_nrtw = 0, desired_nrtw;
	574
cfb185a1	575	/* deal with chardevs and other special file */
	576	if (!mapping->a_ops->writepage)
	577	return 0;
	578
eb47b800 JK	579	if (wbc->nr_to_write < MAX_DESIRED_PAGES_WP) {
	580	desired_nrtw = MAX_DESIRED_PAGES_WP;
	581	excess_nrtw = desired_nrtw - wbc->nr_to_write;
	582	wbc->nr_to_write = desired_nrtw;
	583	}
	584
	585	if (!S_ISDIR(inode->i_mode))
	586	mutex_lock(&sbi->writepages);
fa9150a8	587	ret = write_cache_pages(mapping, wbc, __f2fs_writepage, mapping);
eb47b800 JK	588	if (!S_ISDIR(inode->i_mode))
	589	mutex_unlock(&sbi->writepages);
	590	f2fs_submit_bio(sbi, DATA, (wbc->sync_mode == WB_SYNC_ALL));
	591
	592	remove_dirty_dir_inode(inode);
	593
	594	wbc->nr_to_write -= excess_nrtw;
	595	return ret;
	596	}
	597
	598	static int f2fs_write_begin(struct file file, struct address_space mapping,
	599	loff_t pos, unsigned len, unsigned flags,
	600	struct page pagep, void fsdata)
	601	{
	602	struct inode *inode = mapping->host;
	603	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	604	struct page *page;
	605	pgoff_t index = ((unsigned long long) pos) >> PAGE_CACHE_SHIFT;
	606	struct dnode_of_data dn;
	607	int err = 0;
39936837	608	int ilock;
eb47b800 JK	609
	610	/* for nobh_write_end */
	611	*fsdata = NULL;
	612
	613	f2fs_balance_fs(sbi);
	614
	615	page = grab_cache_page_write_begin(mapping, index, flags);
	616	if (!page)
	617	return -ENOMEM;
	618	*pagep = page;
	619
39936837	620	ilock = mutex_lock_op(sbi);
eb47b800 JK	621
eb47b800 JK	622	set_new_dnode(&dn, inode, NULL, NULL, 0);
266e97a8	623	err = get_dnode_of_data(&dn, index, ALLOC_NODE);
39936837 JK	624	if (err)
39936837 JK	625	goto err;
eb47b800	626
39936837	627	if (dn.data_blkaddr == NULL_ADDR)
eb47b800	628	err = reserve_new_block(&dn);
39936837	629
eb47b800	630	f2fs_put_dnode(&dn);
39936837 JK	631	if (err)
39936837 JK	632	goto err;
eb47b800	633
39936837	634	mutex_unlock_op(sbi, ilock);
eb47b800 JK	635
	636	if ((len == PAGE_CACHE_SIZE) \|\| PageUptodate(page))
	637	return 0;
	638
	639	if ((pos & PAGE_CACHE_MASK) >= i_size_read(inode)) {
	640	unsigned start = pos & (PAGE_CACHE_SIZE - 1);
	641	unsigned end = start + len;
	642
	643	/* Reading beyond i_size is simple: memset to zero */
	644	zero_user_segments(page, 0, start, end, PAGE_CACHE_SIZE);
393ff91f	645	goto out;
eb47b800 JK	646	}
	647
	648	if (dn.data_blkaddr == NEW_ADDR) {
	649	zero_user_segment(page, 0, PAGE_CACHE_SIZE);
	650	} else {
	651	err = f2fs_readpage(sbi, page, dn.data_blkaddr, READ_SYNC);
393ff91f	652	if (err)
eb47b800	653	return err;
393ff91f JK	654	lock_page(page);
	655	if (!PageUptodate(page)) {
	656	f2fs_put_page(page, 1);
	657	return -EIO;
eb47b800 JK	658	}
eb47b800 JK	659	}
393ff91f	660	out:
eb47b800 JK	661	SetPageUptodate(page);
	662	clear_cold_data(page);
	663	return 0;
39936837 JK	664
	665	err:
	666	mutex_unlock_op(sbi, ilock);
	667	f2fs_put_page(page, 1);
	668	return err;
eb47b800 JK	669	}
	670
	671	static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb,
	672	const struct iovec *iov, loff_t offset, unsigned long nr_segs)
	673	{
	674	struct file *file = iocb->ki_filp;
	675	struct inode *inode = file->f_mapping->host;
	676
	677	if (rw == WRITE)
	678	return 0;
	679
	680	/* Needs synchronization with the cleaner */
	681	return blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
	682	get_data_block_ro);
	683	}
	684
	685	static void f2fs_invalidate_data_page(struct page *page, unsigned long offset)
	686	{
	687	struct inode *inode = page->mapping->host;
	688	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
	689	if (S_ISDIR(inode->i_mode) && PageDirty(page)) {
	690	dec_page_count(sbi, F2FS_DIRTY_DENTS);
	691	inode_dec_dirty_dents(inode);
	692	}
	693	ClearPagePrivate(page);
	694	}
	695
	696	static int f2fs_release_data_page(struct page *page, gfp_t wait)
	697	{
	698	ClearPagePrivate(page);
c3850aa1	699	return 1;
eb47b800 JK	700	}
	701
	702	static int f2fs_set_data_page_dirty(struct page *page)
	703	{
	704	struct address_space *mapping = page->mapping;
	705	struct inode *inode = mapping->host;
	706
	707	SetPageUptodate(page);
	708	if (!PageDirty(page)) {
	709	__set_page_dirty_nobuffers(page);
	710	set_dirty_dir_page(inode, page);
	711	return 1;
	712	}
	713	return 0;
	714	}
	715
c01e54b7 JK	716	static sector_t f2fs_bmap(struct address_space *mapping, sector_t block)
	717	{
	718	return generic_block_bmap(mapping, block, get_data_block_ro);
	719	}
	720
eb47b800 JK	721	const struct address_space_operations f2fs_dblock_aops = {
	722	.readpage = f2fs_read_data_page,
	723	.readpages = f2fs_read_data_pages,
	724	.writepage = f2fs_write_data_page,
	725	.writepages = f2fs_write_data_pages,
	726	.write_begin = f2fs_write_begin,
	727	.write_end = nobh_write_end,
	728	.set_page_dirty = f2fs_set_data_page_dirty,
	729	.invalidatepage = f2fs_invalidate_data_page,
	730	.releasepage = f2fs_release_data_page,
	731	.direct_IO = f2fs_direct_IO,
c01e54b7	732	.bmap = f2fs_bmap,
eb47b800	733	};