[mirror_ubuntu-jammy-kernel.git] / fs / nilfs2 / page.c

// SPDX-License-Identifier: GPL-2.0+
/*
 * page.c - buffer/page management specific to NILFS
 *
 * Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation.
 *
 * Written by Ryusuke Konishi and Seiji Kihara.
 */

#include <linux/pagemap.h>
#include <linux/writeback.h>
#include <linux/swap.h>
#include <linux/bitops.h>
#include <linux/page-flags.h>
#include <linux/list.h>
#include <linux/highmem.h>
#include <linux/pagevec.h>
#include <linux/gfp.h>
#include "nilfs.h"
#include "page.h"
#include "mdt.h"


#define NILFS_BUFFER_INHERENT_BITS					\
	(BIT(BH_Uptodate) | BIT(BH_Mapped) | BIT(BH_NILFS_Node) |	\
	 BIT(BH_NILFS_Volatile) | BIT(BH_NILFS_Checked))

static struct buffer_head *
__nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index,
		       int blkbits, unsigned long b_state)

{
	unsigned long first_block;
	struct buffer_head *bh;

	if (!page_has_buffers(page))
		create_empty_buffers(page, 1 << blkbits, b_state);

	first_block = (unsigned long)index << (PAGE_SHIFT - blkbits);
	bh = nilfs_page_get_nth_block(page, block - first_block);

	touch_buffer(bh);
	wait_on_buffer(bh);
	return bh;
}

struct buffer_head *nilfs_grab_buffer(struct inode *inode,
				      struct address_space *mapping,
				      unsigned long blkoff,
				      unsigned long b_state)
{
	int blkbits = inode->i_blkbits;
	pgoff_t index = blkoff >> (PAGE_SHIFT - blkbits);
	struct page *page;
	struct buffer_head *bh;

	page = grab_cache_page(mapping, index);
	if (unlikely(!page))
		return NULL;

	bh = __nilfs_get_page_block(page, blkoff, index, blkbits, b_state);
	if (unlikely(!bh)) {
		unlock_page(page);
		put_page(page);
		return NULL;
	}
	return bh;
}

/**
 * nilfs_forget_buffer - discard dirty state
 * @inode: owner inode of the buffer
 * @bh: buffer head of the buffer to be discarded
 */
void nilfs_forget_buffer(struct buffer_head *bh)
{
	struct page *page = bh->b_page;
	const unsigned long clear_bits =
		(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
		 BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
		 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));

	lock_buffer(bh);
	set_mask_bits(&bh->b_state, clear_bits, 0);
	if (nilfs_page_buffers_clean(page))
		__nilfs_clear_page_dirty(page);

	bh->b_blocknr = -1;
	ClearPageUptodate(page);
	ClearPageMappedToDisk(page);
	unlock_buffer(bh);
	brelse(bh);
}

/**
 * nilfs_copy_buffer -- copy buffer data and flags
 * @dbh: destination buffer
 * @sbh: source buffer
 */
void nilfs_copy_buffer(struct buffer_head *dbh, struct buffer_head *sbh)
{
	void *kaddr0, *kaddr1;
	unsigned long bits;
	struct page *spage = sbh->b_page, *dpage = dbh->b_page;
	struct buffer_head *bh;

	kaddr0 = kmap_atomic(spage);
	kaddr1 = kmap_atomic(dpage);
	memcpy(kaddr1 + bh_offset(dbh), kaddr0 + bh_offset(sbh), sbh->b_size);
	kunmap_atomic(kaddr1);
	kunmap_atomic(kaddr0);

	dbh->b_state = sbh->b_state & NILFS_BUFFER_INHERENT_BITS;
	dbh->b_blocknr = sbh->b_blocknr;
	dbh->b_bdev = sbh->b_bdev;

	bh = dbh;
	bits = sbh->b_state & (BIT(BH_Uptodate) | BIT(BH_Mapped));
	while ((bh = bh->b_this_page) != dbh) {
		lock_buffer(bh);
		bits &= bh->b_state;
		unlock_buffer(bh);
	}
	if (bits & BIT(BH_Uptodate))
		SetPageUptodate(dpage);
	else
		ClearPageUptodate(dpage);
	if (bits & BIT(BH_Mapped))
		SetPageMappedToDisk(dpage);
	else
		ClearPageMappedToDisk(dpage);
}

/**
 * nilfs_page_buffers_clean - check if a page has dirty buffers or not.
 * @page: page to be checked
 *
 * nilfs_page_buffers_clean() returns zero if the page has dirty buffers.
 * Otherwise, it returns non-zero value.
 */
int nilfs_page_buffers_clean(struct page *page)
{
	struct buffer_head *bh, *head;

	bh = head = page_buffers(page);
	do {
		if (buffer_dirty(bh))
			return 0;
		bh = bh->b_this_page;
	} while (bh != head);
	return 1;
}

void nilfs_page_bug(struct page *page)
{
	struct address_space *m;
	unsigned long ino;

	if (unlikely(!page)) {
		printk(KERN_CRIT "NILFS_PAGE_BUG(NULL)\n");
		return;
	}

	m = page->mapping;
	ino = m ? m->host->i_ino : 0;

	printk(KERN_CRIT "NILFS_PAGE_BUG(%p): cnt=%d index#=%llu flags=0x%lx "
	       "mapping=%p ino=%lu\n",
	       page, page_ref_count(page),
	       (unsigned long long)page->index, page->flags, m, ino);

	if (page_has_buffers(page)) {
		struct buffer_head *bh, *head;
		int i = 0;

		bh = head = page_buffers(page);
		do {
			printk(KERN_CRIT
			       " BH[%d] %p: cnt=%d block#=%llu state=0x%lx\n",
			       i++, bh, atomic_read(&bh->b_count),
			       (unsigned long long)bh->b_blocknr, bh->b_state);
			bh = bh->b_this_page;
		} while (bh != head);
	}
}

/**
 * nilfs_copy_page -- copy the page with buffers
 * @dst: destination page
 * @src: source page
 * @copy_dirty: flag whether to copy dirty states on the page's buffer heads.
 *
 * This function is for both data pages and btnode pages.  The dirty flag
 * should be treated by caller.  The page must not be under i/o.
 * Both src and dst page must be locked
 */
static void nilfs_copy_page(struct page *dst, struct page *src, int copy_dirty)
{
	struct buffer_head *dbh, *dbufs, *sbh, *sbufs;
	unsigned long mask = NILFS_BUFFER_INHERENT_BITS;

	BUG_ON(PageWriteback(dst));

	sbh = sbufs = page_buffers(src);
	if (!page_has_buffers(dst))
		create_empty_buffers(dst, sbh->b_size, 0);

	if (copy_dirty)
		mask |= BIT(BH_Dirty);

	dbh = dbufs = page_buffers(dst);
	do {
		lock_buffer(sbh);
		lock_buffer(dbh);
		dbh->b_state = sbh->b_state & mask;
		dbh->b_blocknr = sbh->b_blocknr;
		dbh->b_bdev = sbh->b_bdev;
		sbh = sbh->b_this_page;
		dbh = dbh->b_this_page;
	} while (dbh != dbufs);

	copy_highpage(dst, src);

	if (PageUptodate(src) && !PageUptodate(dst))
		SetPageUptodate(dst);
	else if (!PageUptodate(src) && PageUptodate(dst))
		ClearPageUptodate(dst);
	if (PageMappedToDisk(src) && !PageMappedToDisk(dst))
		SetPageMappedToDisk(dst);
	else if (!PageMappedToDisk(src) && PageMappedToDisk(dst))
		ClearPageMappedToDisk(dst);

	do {
		unlock_buffer(sbh);
		unlock_buffer(dbh);
		sbh = sbh->b_this_page;
		dbh = dbh->b_this_page;
	} while (dbh != dbufs);
}

int nilfs_copy_dirty_pages(struct address_space *dmap,
			   struct address_space *smap)
{
	struct pagevec pvec;
	unsigned int i;
	pgoff_t index = 0;
	int err = 0;

	pagevec_init(&pvec);
repeat:
	if (!pagevec_lookup_tag(&pvec, smap, &index, PAGECACHE_TAG_DIRTY))
		return 0;

	for (i = 0; i < pagevec_count(&pvec); i++) {
		struct page *page = pvec.pages[i], *dpage;

		lock_page(page);
		if (unlikely(!PageDirty(page)))
			NILFS_PAGE_BUG(page, "inconsistent dirty state");

		dpage = grab_cache_page(dmap, page->index);
		if (unlikely(!dpage)) {
			/* No empty page is added to the page cache */
			err = -ENOMEM;
			unlock_page(page);
			break;
		}
		if (unlikely(!page_has_buffers(page)))
			NILFS_PAGE_BUG(page,
				       "found empty page in dat page cache");

		nilfs_copy_page(dpage, page, 1);
		__set_page_dirty_nobuffers(dpage);

		unlock_page(dpage);
		put_page(dpage);
		unlock_page(page);
	}
	pagevec_release(&pvec);
	cond_resched();

	if (likely(!err))
		goto repeat;
	return err;
}

/**
 * nilfs_copy_back_pages -- copy back pages to original cache from shadow cache
 * @dmap: destination page cache
 * @smap: source page cache
 *
 * No pages must no be added to the cache during this process.
 * This must be ensured by the caller.
 */
void nilfs_copy_back_pages(struct address_space *dmap,
			   struct address_space *smap)
{
	struct pagevec pvec;
	unsigned int i, n;
	pgoff_t index = 0;
	int err;

	pagevec_init(&pvec);
repeat:
	n = pagevec_lookup(&pvec, smap, &index);
	if (!n)
		return;

	for (i = 0; i < pagevec_count(&pvec); i++) {
		struct page *page = pvec.pages[i], *dpage;
		pgoff_t offset = page->index;

		lock_page(page);
		dpage = find_lock_page(dmap, offset);
		if (dpage) {
			/* override existing page on the destination cache */
			WARN_ON(PageDirty(dpage));
			nilfs_copy_page(dpage, page, 0);
			unlock_page(dpage);
			put_page(dpage);
		} else {
			struct page *page2;

			/* move the page to the destination cache */
			xa_lock_irq(&smap->i_pages);
			page2 = radix_tree_delete(&smap->i_pages, offset);
			WARN_ON(page2 != page);

			smap->nrpages--;
			xa_unlock_irq(&smap->i_pages);

			xa_lock_irq(&dmap->i_pages);
			err = radix_tree_insert(&dmap->i_pages, offset, page);
			if (unlikely(err < 0)) {
				WARN_ON(err == -EEXIST);
				page->mapping = NULL;
				put_page(page); /* for cache */
			} else {
				page->mapping = dmap;
				dmap->nrpages++;
				if (PageDirty(page))
					radix_tree_tag_set(&dmap->i_pages,
							   offset,
							   PAGECACHE_TAG_DIRTY);
			}
			xa_unlock_irq(&dmap->i_pages);
		}
		unlock_page(page);
	}
	pagevec_release(&pvec);
	cond_resched();

	goto repeat;
}

/**
 * nilfs_clear_dirty_pages - discard dirty pages in address space
 * @mapping: address space with dirty pages for discarding
 * @silent: suppress [true] or print [false] warning messages
 */
void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
{
	struct pagevec pvec;
	unsigned int i;
	pgoff_t index = 0;

	pagevec_init(&pvec);

	while (pagevec_lookup_tag(&pvec, mapping, &index,
					PAGECACHE_TAG_DIRTY)) {
		for (i = 0; i < pagevec_count(&pvec); i++) {
			struct page *page = pvec.pages[i];

			lock_page(page);
			nilfs_clear_dirty_page(page, silent);
			unlock_page(page);
		}
		pagevec_release(&pvec);
		cond_resched();
	}
}

/**
 * nilfs_clear_dirty_page - discard dirty page
 * @page: dirty page that will be discarded
 * @silent: suppress [true] or print [false] warning messages
 */
void nilfs_clear_dirty_page(struct page *page, bool silent)
{
	struct inode *inode = page->mapping->host;
	struct super_block *sb = inode->i_sb;

	BUG_ON(!PageLocked(page));

	if (!silent)
		nilfs_msg(sb, KERN_WARNING,
			  "discard dirty page: offset=%lld, ino=%lu",
			  page_offset(page), inode->i_ino);

	ClearPageUptodate(page);
	ClearPageMappedToDisk(page);

	if (page_has_buffers(page)) {
		struct buffer_head *bh, *head;
		const unsigned long clear_bits =
			(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
			 BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
			 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));

		bh = head = page_buffers(page);
		do {
			lock_buffer(bh);
			if (!silent)
				nilfs_msg(sb, KERN_WARNING,
					  "discard dirty block: blocknr=%llu, size=%zu",
					  (u64)bh->b_blocknr, bh->b_size);

			set_mask_bits(&bh->b_state, clear_bits, 0);
			unlock_buffer(bh);
		} while (bh = bh->b_this_page, bh != head);
	}

	__nilfs_clear_page_dirty(page);
}

unsigned int nilfs_page_count_clean_buffers(struct page *page,
					    unsigned int from, unsigned int to)
{
	unsigned int block_start, block_end;
	struct buffer_head *bh, *head;
	unsigned int nc = 0;

	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 + bh->b_size;
		if (block_end > from && block_start < to && !buffer_dirty(bh))
			nc++;
	}
	return nc;
}

void nilfs_mapping_init(struct address_space *mapping, struct inode *inode)
{
	mapping->host = inode;
	mapping->flags = 0;
	mapping_set_gfp_mask(mapping, GFP_NOFS);
	mapping->private_data = NULL;
	mapping->a_ops = &empty_aops;
}

/*
 * NILFS2 needs clear_page_dirty() in the following two cases:
 *
 * 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears
 *    page dirty flags when it copies back pages from the shadow cache
 *    (gcdat->{i_mapping,i_btnode_cache}) to its original cache
 *    (dat->{i_mapping,i_btnode_cache}).
 *
 * 2) Some B-tree operations like insertion or deletion may dispose buffers
 *    in dirty state, and this needs to cancel the dirty state of their pages.
 */
int __nilfs_clear_page_dirty(struct page *page)
{
	struct address_space *mapping = page->mapping;

	if (mapping) {
		xa_lock_irq(&mapping->i_pages);
		if (test_bit(PG_dirty, &page->flags)) {
			radix_tree_tag_clear(&mapping->i_pages,
					     page_index(page),
					     PAGECACHE_TAG_DIRTY);
			xa_unlock_irq(&mapping->i_pages);
			return clear_page_dirty_for_io(page);
		}
		xa_unlock_irq(&mapping->i_pages);
		return 0;
	}
	return TestClearPageDirty(page);
}

/**
 * nilfs_find_uncommitted_extent - find extent of uncommitted data
 * @inode: inode
 * @start_blk: start block offset (in)
 * @blkoff: start offset of the found extent (out)
 *
 * This function searches an extent of buffers marked "delayed" which
 * starts from a block offset equal to or larger than @start_blk.  If
 * such an extent was found, this will store the start offset in
 * @blkoff and return its length in blocks.  Otherwise, zero is
 * returned.
 */
unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
					    sector_t start_blk,
					    sector_t *blkoff)
{
	unsigned int i;
	pgoff_t index;
	unsigned int nblocks_in_page;
	unsigned long length = 0;
	sector_t b;
	struct pagevec pvec;
	struct page *page;

	if (inode->i_mapping->nrpages == 0)
		return 0;

	index = start_blk >> (PAGE_SHIFT - inode->i_blkbits);
	nblocks_in_page = 1U << (PAGE_SHIFT - inode->i_blkbits);

	pagevec_init(&pvec);

repeat:
	pvec.nr = find_get_pages_contig(inode->i_mapping, index, PAGEVEC_SIZE,
					pvec.pages);
	if (pvec.nr == 0)
		return length;

	if (length > 0 && pvec.pages[0]->index > index)
		goto out;

	b = pvec.pages[0]->index << (PAGE_SHIFT - inode->i_blkbits);
	i = 0;
	do {
		page = pvec.pages[i];

		lock_page(page);
		if (page_has_buffers(page)) {
			struct buffer_head *bh, *head;

			bh = head = page_buffers(page);
			do {
				if (b < start_blk)
					continue;
				if (buffer_delay(bh)) {
					if (length == 0)
						*blkoff = b;
					length++;
				} else if (length > 0) {
					goto out_locked;
				}
			} while (++b, bh = bh->b_this_page, bh != head);
		} else {
			if (length > 0)
				goto out_locked;

			b += nblocks_in_page;
		}
		unlock_page(page);

	} while (++i < pagevec_count(&pvec));

	index = page->index + 1;
	pagevec_release(&pvec);
	cond_resched();
	goto repeat;

out_locked:
	unlock_page(page);
out:
	pagevec_release(&pvec);
	return length;
}
Commit	Line	Data
ae98043f	1	// SPDX-License-Identifier: GPL-2.0+
0bd49f94 RK	2	/*
	3	* page.c - buffer/page management specific to NILFS
	4	*
	5	* Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation.
	6	*
4b420ab4	7	* Written by Ryusuke Konishi and Seiji Kihara.
0bd49f94 RK	8	*/
	9
	10	#include <linux/pagemap.h>
	11	#include <linux/writeback.h>
	12	#include <linux/swap.h>
	13	#include <linux/bitops.h>
	14	#include <linux/page-flags.h>
	15	#include <linux/list.h>
	16	#include <linux/highmem.h>
	17	#include <linux/pagevec.h>
5a0e3ad6	18	#include <linux/gfp.h>
0bd49f94 RK	19	#include "nilfs.h"
	20	#include "page.h"
	21	#include "mdt.h"
	22
	23
4ce5c342 RK	24	#define NILFS_BUFFER_INHERENT_BITS \
	25	(BIT(BH_Uptodate) \| BIT(BH_Mapped) \| BIT(BH_NILFS_Node) \| \
	26	BIT(BH_NILFS_Volatile) \| BIT(BH_NILFS_Checked))
0bd49f94 RK	27
	28	static struct buffer_head *
	29	__nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index,
	30	int blkbits, unsigned long b_state)
	31
	32	{
	33	unsigned long first_block;
	34	struct buffer_head *bh;
	35
	36	if (!page_has_buffers(page))
	37	create_empty_buffers(page, 1 << blkbits, b_state);
	38
09cbfeaf	39	first_block = (unsigned long)index << (PAGE_SHIFT - blkbits);
0bd49f94 RK	40	bh = nilfs_page_get_nth_block(page, block - first_block);
	41
	42	touch_buffer(bh);
	43	wait_on_buffer(bh);
	44	return bh;
	45	}
	46
0bd49f94 RK	47	struct buffer_head nilfs_grab_buffer(struct inode inode,
	48	struct address_space *mapping,
	49	unsigned long blkoff,
	50	unsigned long b_state)
	51	{
	52	int blkbits = inode->i_blkbits;
09cbfeaf	53	pgoff_t index = blkoff >> (PAGE_SHIFT - blkbits);
c1c1d709 RK	54	struct page *page;
c1c1d709 RK	55	struct buffer_head *bh;
0bd49f94 RK	56
	57	page = grab_cache_page(mapping, index);
	58	if (unlikely(!page))
	59	return NULL;
	60
	61	bh = __nilfs_get_page_block(page, blkoff, index, blkbits, b_state);
	62	if (unlikely(!bh)) {
	63	unlock_page(page);
09cbfeaf	64	put_page(page);
0bd49f94 RK	65	return NULL;
0bd49f94 RK	66	}
0bd49f94 RK	67	return bh;
	68	}
	69
	70	/**
	71	* nilfs_forget_buffer - discard dirty state
	72	* @inode: owner inode of the buffer
	73	* @bh: buffer head of the buffer to be discarded
	74	*/
	75	void nilfs_forget_buffer(struct buffer_head *bh)
	76	{
	77	struct page *page = bh->b_page;
ead8ecff	78	const unsigned long clear_bits =
4ce5c342 RK	79	(BIT(BH_Uptodate) \| BIT(BH_Dirty) \| BIT(BH_Mapped) \|
	80	BIT(BH_Async_Write) \| BIT(BH_NILFS_Volatile) \|
	81	BIT(BH_NILFS_Checked) \| BIT(BH_NILFS_Redirected));
0bd49f94 RK	82
0bd49f94 RK	83	lock_buffer(bh);
ead8ecff	84	set_mask_bits(&bh->b_state, clear_bits, 0);
84338237	85	if (nilfs_page_buffers_clean(page))
0bd49f94 RK	86	__nilfs_clear_page_dirty(page);
0bd49f94 RK	87
0bd49f94 RK	88	bh->b_blocknr = -1;
	89	ClearPageUptodate(page);
	90	ClearPageMappedToDisk(page);
	91	unlock_buffer(bh);
	92	brelse(bh);
	93	}
	94
	95	/**
	96	* nilfs_copy_buffer -- copy buffer data and flags
	97	* @dbh: destination buffer
	98	* @sbh: source buffer
	99	*/
	100	void nilfs_copy_buffer(struct buffer_head dbh, struct buffer_head sbh)
	101	{
	102	void kaddr0, kaddr1;
	103	unsigned long bits;
	104	struct page spage = sbh->b_page, dpage = dbh->b_page;
	105	struct buffer_head *bh;
	106
7b9c0976 CW	107	kaddr0 = kmap_atomic(spage);
7b9c0976 CW	108	kaddr1 = kmap_atomic(dpage);
0bd49f94	109	memcpy(kaddr1 + bh_offset(dbh), kaddr0 + bh_offset(sbh), sbh->b_size);
7b9c0976 CW	110	kunmap_atomic(kaddr1);
7b9c0976 CW	111	kunmap_atomic(kaddr0);
0bd49f94 RK	112
	113	dbh->b_state = sbh->b_state & NILFS_BUFFER_INHERENT_BITS;
	114	dbh->b_blocknr = sbh->b_blocknr;
	115	dbh->b_bdev = sbh->b_bdev;
	116
	117	bh = dbh;
4ce5c342	118	bits = sbh->b_state & (BIT(BH_Uptodate) \| BIT(BH_Mapped));
0bd49f94 RK	119	while ((bh = bh->b_this_page) != dbh) {
	120	lock_buffer(bh);
	121	bits &= bh->b_state;
	122	unlock_buffer(bh);
	123	}
4ce5c342	124	if (bits & BIT(BH_Uptodate))
0bd49f94 RK	125	SetPageUptodate(dpage);
	126	else
	127	ClearPageUptodate(dpage);
4ce5c342	128	if (bits & BIT(BH_Mapped))
0bd49f94 RK	129	SetPageMappedToDisk(dpage);
	130	else
	131	ClearPageMappedToDisk(dpage);
	132	}
	133
	134	/**
	135	* nilfs_page_buffers_clean - check if a page has dirty buffers or not.
	136	* @page: page to be checked
	137	*
	138	* nilfs_page_buffers_clean() returns zero if the page has dirty buffers.
	139	* Otherwise, it returns non-zero value.
	140	*/
	141	int nilfs_page_buffers_clean(struct page *page)
	142	{
	143	struct buffer_head bh, head;
	144
	145	bh = head = page_buffers(page);
	146	do {
	147	if (buffer_dirty(bh))
	148	return 0;
	149	bh = bh->b_this_page;
	150	} while (bh != head);
	151	return 1;
	152	}
	153
	154	void nilfs_page_bug(struct page *page)
	155	{
	156	struct address_space *m;
aa405b1f	157	unsigned long ino;
0bd49f94 RK	158
	159	if (unlikely(!page)) {
	160	printk(KERN_CRIT "NILFS_PAGE_BUG(NULL)\n");
	161	return;
	162	}
	163
	164	m = page->mapping;
aa405b1f RK	165	ino = m ? m->host->i_ino : 0;
aa405b1f RK	166
0bd49f94 RK	167	printk(KERN_CRIT "NILFS_PAGE_BUG(%p): cnt=%d index#=%llu flags=0x%lx "
0bd49f94 RK	168	"mapping=%p ino=%lu\n",
fe896d18	169	page, page_ref_count(page),
0bd49f94 RK	170	(unsigned long long)page->index, page->flags, m, ino);
	171
	172	if (page_has_buffers(page)) {
	173	struct buffer_head bh, head;
	174	int i = 0;
	175
	176	bh = head = page_buffers(page);
	177	do {
	178	printk(KERN_CRIT
	179	" BH[%d] %p: cnt=%d block#=%llu state=0x%lx\n",
	180	i++, bh, atomic_read(&bh->b_count),
	181	(unsigned long long)bh->b_blocknr, bh->b_state);
	182	bh = bh->b_this_page;
	183	} while (bh != head);
	184	}
	185	}
	186
0bd49f94 RK	187	/**
	188	* nilfs_copy_page -- copy the page with buffers
	189	* @dst: destination page
	190	* @src: source page
	191	* @copy_dirty: flag whether to copy dirty states on the page's buffer heads.
	192	*
7a65004b	193	* This function is for both data pages and btnode pages. The dirty flag
0bd49f94 RK	194	* should be treated by caller. The page must not be under i/o.
	195	* Both src and dst page must be locked
	196	*/
	197	static void nilfs_copy_page(struct page dst, struct page src, int copy_dirty)
	198	{
	199	struct buffer_head dbh, dbufs, sbh, sbufs;
	200	unsigned long mask = NILFS_BUFFER_INHERENT_BITS;
	201
	202	BUG_ON(PageWriteback(dst));
	203
	204	sbh = sbufs = page_buffers(src);
	205	if (!page_has_buffers(dst))
	206	create_empty_buffers(dst, sbh->b_size, 0);
	207
	208	if (copy_dirty)
4ce5c342	209	mask \|= BIT(BH_Dirty);
0bd49f94 RK	210
	211	dbh = dbufs = page_buffers(dst);
	212	do {
	213	lock_buffer(sbh);
	214	lock_buffer(dbh);
	215	dbh->b_state = sbh->b_state & mask;
	216	dbh->b_blocknr = sbh->b_blocknr;
	217	dbh->b_bdev = sbh->b_bdev;
	218	sbh = sbh->b_this_page;
	219	dbh = dbh->b_this_page;
	220	} while (dbh != dbufs);
	221
	222	copy_highpage(dst, src);
	223
	224	if (PageUptodate(src) && !PageUptodate(dst))
	225	SetPageUptodate(dst);
	226	else if (!PageUptodate(src) && PageUptodate(dst))
	227	ClearPageUptodate(dst);
	228	if (PageMappedToDisk(src) && !PageMappedToDisk(dst))
	229	SetPageMappedToDisk(dst);
	230	else if (!PageMappedToDisk(src) && PageMappedToDisk(dst))
	231	ClearPageMappedToDisk(dst);
	232
	233	do {
	234	unlock_buffer(sbh);
	235	unlock_buffer(dbh);
	236	sbh = sbh->b_this_page;
	237	dbh = dbh->b_this_page;
	238	} while (dbh != dbufs);
	239	}
	240
	241	int nilfs_copy_dirty_pages(struct address_space *dmap,
	242	struct address_space *smap)
	243	{
	244	struct pagevec pvec;
	245	unsigned int i;
	246	pgoff_t index = 0;
	247	int err = 0;
	248
86679820	249	pagevec_init(&pvec);
0bd49f94	250	repeat:
67fd707f	251	if (!pagevec_lookup_tag(&pvec, smap, &index, PAGECACHE_TAG_DIRTY))
0bd49f94 RK	252	return 0;
	253
	254	for (i = 0; i < pagevec_count(&pvec); i++) {
	255	struct page page = pvec.pages[i], dpage;
	256
	257	lock_page(page);
	258	if (unlikely(!PageDirty(page)))
	259	NILFS_PAGE_BUG(page, "inconsistent dirty state");
	260
	261	dpage = grab_cache_page(dmap, page->index);
	262	if (unlikely(!dpage)) {
	263	/* No empty page is added to the page cache */
	264	err = -ENOMEM;
	265	unlock_page(page);
	266	break;
	267	}
	268	if (unlikely(!page_has_buffers(page)))
	269	NILFS_PAGE_BUG(page,
	270	"found empty page in dat page cache");
	271
	272	nilfs_copy_page(dpage, page, 1);
	273	__set_page_dirty_nobuffers(dpage);
	274
	275	unlock_page(dpage);
09cbfeaf	276	put_page(dpage);
0bd49f94 RK	277	unlock_page(page);
	278	}
	279	pagevec_release(&pvec);
	280	cond_resched();
	281
	282	if (likely(!err))
	283	goto repeat;
	284	return err;
	285	}
	286
	287	/**
7a65004b	288	* nilfs_copy_back_pages -- copy back pages to original cache from shadow cache
0bd49f94 RK	289	* @dmap: destination page cache
	290	* @smap: source page cache
	291	*
	292	* No pages must no be added to the cache during this process.
	293	* This must be ensured by the caller.
	294	*/
	295	void nilfs_copy_back_pages(struct address_space *dmap,
	296	struct address_space *smap)
	297	{
	298	struct pagevec pvec;
	299	unsigned int i, n;
	300	pgoff_t index = 0;
	301	int err;
	302
86679820	303	pagevec_init(&pvec);
0bd49f94	304	repeat:
397162ff	305	n = pagevec_lookup(&pvec, smap, &index);
0bd49f94 RK	306	if (!n)
0bd49f94 RK	307	return;
0bd49f94 RK	308
	309	for (i = 0; i < pagevec_count(&pvec); i++) {
	310	struct page page = pvec.pages[i], dpage;
	311	pgoff_t offset = page->index;
	312
	313	lock_page(page);
	314	dpage = find_lock_page(dmap, offset);
	315	if (dpage) {
	316	/* override existing page on the destination cache */
1f5abe7e	317	WARN_ON(PageDirty(dpage));
0bd49f94 RK	318	nilfs_copy_page(dpage, page, 0);
0bd49f94 RK	319	unlock_page(dpage);
09cbfeaf	320	put_page(dpage);
0bd49f94 RK	321	} else {
	322	struct page *page2;
	323
	324	/* move the page to the destination cache */
b93b0163 MW	325	xa_lock_irq(&smap->i_pages);
b93b0163 MW	326	page2 = radix_tree_delete(&smap->i_pages, offset);
1f5abe7e RK	327	WARN_ON(page2 != page);
1f5abe7e RK	328
0bd49f94	329	smap->nrpages--;
b93b0163	330	xa_unlock_irq(&smap->i_pages);
0bd49f94	331
b93b0163 MW	332	xa_lock_irq(&dmap->i_pages);
b93b0163 MW	333	err = radix_tree_insert(&dmap->i_pages, offset, page);
0bd49f94	334	if (unlikely(err < 0)) {
1f5abe7e	335	WARN_ON(err == -EEXIST);
0bd49f94	336	page->mapping = NULL;
09cbfeaf	337	put_page(page); /* for cache */
0bd49f94 RK	338	} else {
	339	page->mapping = dmap;
	340	dmap->nrpages++;
	341	if (PageDirty(page))
b93b0163	342	radix_tree_tag_set(&dmap->i_pages,
0bd49f94 RK	343	offset,
	344	PAGECACHE_TAG_DIRTY);
	345	}
b93b0163	346	xa_unlock_irq(&dmap->i_pages);
0bd49f94 RK	347	}
	348	unlock_page(page);
	349	}
	350	pagevec_release(&pvec);
	351	cond_resched();
	352
	353	goto repeat;
	354	}
	355
8c26c4e2 VD	356	/**
	357	* nilfs_clear_dirty_pages - discard dirty pages in address space
	358	* @mapping: address space with dirty pages for discarding
	359	* @silent: suppress [true] or print [false] warning messages
	360	*/
	361	void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
0bd49f94 RK	362	{
	363	struct pagevec pvec;
	364	unsigned int i;
	365	pgoff_t index = 0;
	366
86679820	367	pagevec_init(&pvec);
0bd49f94	368
67fd707f JK	369	while (pagevec_lookup_tag(&pvec, mapping, &index,
67fd707f JK	370	PAGECACHE_TAG_DIRTY)) {
0bd49f94 RK	371	for (i = 0; i < pagevec_count(&pvec); i++) {
0bd49f94 RK	372	struct page *page = pvec.pages[i];
0bd49f94 RK	373
0bd49f94 RK	374	lock_page(page);
8c26c4e2	375	nilfs_clear_dirty_page(page, silent);
0bd49f94 RK	376	unlock_page(page);
	377	}
	378	pagevec_release(&pvec);
	379	cond_resched();
	380	}
	381	}
	382
8c26c4e2 VD	383	/**
	384	* nilfs_clear_dirty_page - discard dirty page
	385	* @page: dirty page that will be discarded
	386	* @silent: suppress [true] or print [false] warning messages
	387	*/
	388	void nilfs_clear_dirty_page(struct page *page, bool silent)
	389	{
	390	struct inode *inode = page->mapping->host;
	391	struct super_block *sb = inode->i_sb;
	392
dc33f5f3	393	BUG_ON(!PageLocked(page));
8c26c4e2	394
d6517deb RK	395	if (!silent)
	396	nilfs_msg(sb, KERN_WARNING,
	397	"discard dirty page: offset=%lld, ino=%lu",
	398	page_offset(page), inode->i_ino);
8c26c4e2 VD	399
	400	ClearPageUptodate(page);
	401	ClearPageMappedToDisk(page);
	402
	403	if (page_has_buffers(page)) {
	404	struct buffer_head bh, head;
ead8ecff	405	const unsigned long clear_bits =
4ce5c342 RK	406	(BIT(BH_Uptodate) \| BIT(BH_Dirty) \| BIT(BH_Mapped) \|
	407	BIT(BH_Async_Write) \| BIT(BH_NILFS_Volatile) \|
	408	BIT(BH_NILFS_Checked) \| BIT(BH_NILFS_Redirected));
8c26c4e2 VD	409
	410	bh = head = page_buffers(page);
	411	do {
	412	lock_buffer(bh);
d6517deb RK	413	if (!silent)
	414	nilfs_msg(sb, KERN_WARNING,
	415	"discard dirty block: blocknr=%llu, size=%zu",
	416	(u64)bh->b_blocknr, bh->b_size);
	417
ead8ecff	418	set_mask_bits(&bh->b_state, clear_bits, 0);
8c26c4e2 VD	419	unlock_buffer(bh);
	420	} while (bh = bh->b_this_page, bh != head);
	421	}
	422
	423	__nilfs_clear_page_dirty(page);
	424	}
	425
0c6c44cb RK	426	unsigned int nilfs_page_count_clean_buffers(struct page *page,
0c6c44cb RK	427	unsigned int from, unsigned int to)
0bd49f94	428	{
0c6c44cb	429	unsigned int block_start, block_end;
0bd49f94	430	struct buffer_head bh, head;
0c6c44cb	431	unsigned int nc = 0;
0bd49f94 RK	432
	433	for (bh = head = page_buffers(page), block_start = 0;
	434	bh != head \|\| !block_start;
	435	block_start = block_end, bh = bh->b_this_page) {
	436	block_end = block_start + bh->b_size;
	437	if (block_end > from && block_start < to && !buffer_dirty(bh))
	438	nc++;
	439	}
	440	return nc;
	441	}
ae53a0a2	442
b83ae6d4	443	void nilfs_mapping_init(struct address_space mapping, struct inode inode)
ebdfed4d	444	{
aa405b1f	445	mapping->host = inode;
ebdfed4d RK	446	mapping->flags = 0;
ebdfed4d RK	447	mapping_set_gfp_mask(mapping, GFP_NOFS);
252aa6f5	448	mapping->private_data = NULL;
d611b22f	449	mapping->a_ops = &empty_aops;
ebdfed4d	450	}
0bd49f94 RK	451
	452	/*
	453	* NILFS2 needs clear_page_dirty() in the following two cases:
	454	*
	455	* 1) For B-tree node pages and data pages of the dat/gcdat, NILFS2 clears
	456	* page dirty flags when it copies back pages from the shadow cache
	457	* (gcdat->{i_mapping,i_btnode_cache}) to its original cache
	458	* (dat->{i_mapping,i_btnode_cache}).
	459	*
	460	* 2) Some B-tree operations like insertion or deletion may dispose buffers
	461	* in dirty state, and this needs to cancel the dirty state of their pages.
	462	*/
	463	int __nilfs_clear_page_dirty(struct page *page)
	464	{
	465	struct address_space *mapping = page->mapping;
	466
	467	if (mapping) {
b93b0163	468	xa_lock_irq(&mapping->i_pages);
0bd49f94	469	if (test_bit(PG_dirty, &page->flags)) {
b93b0163	470	radix_tree_tag_clear(&mapping->i_pages,
0bd49f94 RK	471	page_index(page),
0bd49f94 RK	472	PAGECACHE_TAG_DIRTY);
b93b0163	473	xa_unlock_irq(&mapping->i_pages);
0bd49f94 RK	474	return clear_page_dirty_for_io(page);
0bd49f94 RK	475	}
b93b0163	476	xa_unlock_irq(&mapping->i_pages);
0bd49f94 RK	477	return 0;
	478	}
	479	return TestClearPageDirty(page);
	480	}
622daaff RK	481
	482	/**
	483	* nilfs_find_uncommitted_extent - find extent of uncommitted data
	484	* @inode: inode
	485	* @start_blk: start block offset (in)
	486	* @blkoff: start offset of the found extent (out)
	487	*
	488	* This function searches an extent of buffers marked "delayed" which
	489	* starts from a block offset equal to or larger than @start_blk. If
	490	* such an extent was found, this will store the start offset in
	491	* @blkoff and return its length in blocks. Otherwise, zero is
	492	* returned.
	493	*/
	494	unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
	495	sector_t start_blk,
	496	sector_t *blkoff)
	497	{
	498	unsigned int i;
	499	pgoff_t index;
	500	unsigned int nblocks_in_page;
	501	unsigned long length = 0;
	502	sector_t b;
	503	struct pagevec pvec;
	504	struct page *page;
	505
	506	if (inode->i_mapping->nrpages == 0)
	507	return 0;
	508
09cbfeaf KS	509	index = start_blk >> (PAGE_SHIFT - inode->i_blkbits);
09cbfeaf KS	510	nblocks_in_page = 1U << (PAGE_SHIFT - inode->i_blkbits);
622daaff	511
86679820	512	pagevec_init(&pvec);
622daaff RK	513
	514	repeat:
	515	pvec.nr = find_get_pages_contig(inode->i_mapping, index, PAGEVEC_SIZE,
	516	pvec.pages);
	517	if (pvec.nr == 0)
	518	return length;
	519
	520	if (length > 0 && pvec.pages[0]->index > index)
	521	goto out;
	522
09cbfeaf	523	b = pvec.pages[0]->index << (PAGE_SHIFT - inode->i_blkbits);
622daaff RK	524	i = 0;
	525	do {
	526	page = pvec.pages[i];
	527
	528	lock_page(page);
	529	if (page_has_buffers(page)) {
	530	struct buffer_head bh, head;
	531
	532	bh = head = page_buffers(page);
	533	do {
	534	if (b < start_blk)
	535	continue;
	536	if (buffer_delay(bh)) {
	537	if (length == 0)
	538	*blkoff = b;
	539	length++;
	540	} else if (length > 0) {
	541	goto out_locked;
	542	}
	543	} while (++b, bh = bh->b_this_page, bh != head);
	544	} else {
	545	if (length > 0)
	546	goto out_locked;
	547
	548	b += nblocks_in_page;
	549	}
	550	unlock_page(page);
	551
	552	} while (++i < pagevec_count(&pvec));
	553
	554	index = page->index + 1;
	555	pagevec_release(&pvec);
	556	cond_resched();
	557	goto repeat;
	558
	559	out_locked:
	560	unlock_page(page);
	561	out:
	562	pagevec_release(&pvec);
	563	return length;
	564	}