[net-next-2.6.git] / fs / gfs2 / meta_io.c

/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
 * Copyright (C) 2004-2006 Red Hat, Inc.  All rights reserved.
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
 * of the GNU General Public License version 2.
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
#include <linux/mm.h>
#include <linux/pagemap.h>
#include <linux/writeback.h>
#include <linux/swap.h>
#include <linux/delay.h>
#include <linux/bio.h>
#include <linux/gfs2_ondisk.h>
#include <linux/lm_interface.h>

#include "gfs2.h"
#include "incore.h"
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "log.h"
#include "lops.h"
#include "meta_io.h"
#include "rgrp.h"
#include "trans.h"
#include "util.h"
#include "ops_address.h"

static int aspace_get_block(struct inode *inode, sector_t lblock,
			    struct buffer_head *bh_result, int create)
{
	gfs2_assert_warn(inode->i_sb->s_fs_info, 0);
	return -EOPNOTSUPP;
}

static int gfs2_aspace_writepage(struct page *page,
				 struct writeback_control *wbc)
{
	return block_write_full_page(page, aspace_get_block, wbc);
}

static const struct address_space_operations aspace_aops = {
	.writepage = gfs2_aspace_writepage,
	.releasepage = gfs2_releasepage,
};

/**
 * gfs2_aspace_get - Create and initialize a struct inode structure
 * @sdp: the filesystem the aspace is in
 *
 * Right now a struct inode is just a struct inode.  Maybe Linux
 * will supply a more lightweight address space construct (that works)
 * in the future.
 *
 * Make sure pages/buffers in this aspace aren't in high memory.
 *
 * Returns: the aspace
 */

struct inode *gfs2_aspace_get(struct gfs2_sbd *sdp)
{
	struct inode *aspace;

	aspace = new_inode(sdp->sd_vfs);
	if (aspace) {
		mapping_set_gfp_mask(aspace->i_mapping, GFP_NOFS);
		aspace->i_mapping->a_ops = &aspace_aops;
		aspace->i_size = ~0ULL;
		aspace->i_private = NULL;
		insert_inode_hash(aspace);
	}
	return aspace;
}

void gfs2_aspace_put(struct inode *aspace)
{
	remove_inode_hash(aspace);
	iput(aspace);
}

/**
 * gfs2_meta_inval - Invalidate all buffers associated with a glock
 * @gl: the glock
 *
 */

void gfs2_meta_inval(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct inode *aspace = gl->gl_aspace;
	struct address_space *mapping = gl->gl_aspace->i_mapping;

	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));

	atomic_inc(&aspace->i_writecount);
	truncate_inode_pages(mapping, 0);
	atomic_dec(&aspace->i_writecount);

	gfs2_assert_withdraw(sdp, !mapping->nrpages);
}

/**
 * gfs2_meta_sync - Sync all buffers associated with a glock
 * @gl: The glock
 *
 */

void gfs2_meta_sync(struct gfs2_glock *gl)
{
	struct address_space *mapping = gl->gl_aspace->i_mapping;
	int error;

	filemap_fdatawrite(mapping);
	error = filemap_fdatawait(mapping);

	if (error)
		gfs2_io_error(gl->gl_sbd);
}

/**
 * getbuf - Get a buffer with a given address space
 * @gl: the glock
 * @blkno: the block number (filesystem scope)
 * @create: 1 if the buffer should be created
 *
 * Returns: the buffer
 */

static struct buffer_head *getbuf(struct gfs2_glock *gl, u64 blkno, int create)
{
	struct address_space *mapping = gl->gl_aspace->i_mapping;
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct page *page;
	struct buffer_head *bh;
	unsigned int shift;
	unsigned long index;
	unsigned int bufnum;

	shift = PAGE_CACHE_SHIFT - sdp->sd_sb.sb_bsize_shift;
	index = blkno >> shift;             /* convert block to page */
	bufnum = blkno - (index << shift);  /* block buf index within page */

	if (create) {
		for (;;) {
			page = grab_cache_page(mapping, index);
			if (page)
				break;
			yield();
		}
	} else {
		page = find_lock_page(mapping, index);
		if (!page)
			return NULL;
	}

	if (!page_has_buffers(page))
		create_empty_buffers(page, sdp->sd_sb.sb_bsize, 0);

	/* Locate header for our buffer within our page */
	for (bh = page_buffers(page); bufnum--; bh = bh->b_this_page)
		/* Do nothing */;
	get_bh(bh);

	if (!buffer_mapped(bh))
		map_bh(bh, sdp->sd_vfs, blkno);

	unlock_page(page);
	mark_page_accessed(page);
	page_cache_release(page);

	return bh;
}

static void meta_prep_new(struct buffer_head *bh)
{
	struct gfs2_meta_header *mh = (struct gfs2_meta_header *)bh->b_data;

	lock_buffer(bh);
	clear_buffer_dirty(bh);
	set_buffer_uptodate(bh);
	unlock_buffer(bh);

	mh->mh_magic = cpu_to_be32(GFS2_MAGIC);
}

/**
 * gfs2_meta_new - Get a block
 * @gl: The glock associated with this block
 * @blkno: The block number
 *
 * Returns: The buffer
 */

struct buffer_head *gfs2_meta_new(struct gfs2_glock *gl, u64 blkno)
{
	struct buffer_head *bh;
	bh = getbuf(gl, blkno, CREATE);
	meta_prep_new(bh);
	return bh;
}

/**
 * gfs2_meta_read - Read a block from disk
 * @gl: The glock covering the block
 * @blkno: The block number
 * @flags: flags
 * @bhp: the place where the buffer is returned (NULL on failure)
 *
 * Returns: errno
 */

int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
		   struct buffer_head **bhp)
{
	*bhp = getbuf(gl, blkno, CREATE);
	if (!buffer_uptodate(*bhp))
		ll_rw_block(READ_META, 1, bhp);
	if (flags & DIO_WAIT) {
		int error = gfs2_meta_wait(gl->gl_sbd, *bhp);
		if (error) {
			brelse(*bhp);
			return error;
		}
	}

	return 0;
}

/**
 * gfs2_meta_wait - Reread a block from disk
 * @sdp: the filesystem
 * @bh: The block to wait for
 *
 * Returns: errno
 */

int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh)
{
	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
		return -EIO;

	wait_on_buffer(bh);

	if (!buffer_uptodate(bh)) {
		struct gfs2_trans *tr = current->journal_info;
		if (tr && tr->tr_touched)
			gfs2_io_error_bh(sdp, bh);
		return -EIO;
	}
	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
		return -EIO;

	return 0;
}

/**
 * gfs2_attach_bufdata - attach a struct gfs2_bufdata structure to a buffer
 * @gl: the glock the buffer belongs to
 * @bh: The buffer to be attached to
 * @meta: Flag to indicate whether its metadata or not
 */

void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh,
			 int meta)
{
	struct gfs2_bufdata *bd;

	if (meta)
		lock_page(bh->b_page);

	if (bh->b_private) {
		if (meta)
			unlock_page(bh->b_page);
		return;
	}

	bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL),
	bd->bd_bh = bh;
	bd->bd_gl = gl;

	INIT_LIST_HEAD(&bd->bd_list_tr);
	if (meta)
		lops_init_le(&bd->bd_le, &gfs2_buf_lops);
	else
		lops_init_le(&bd->bd_le, &gfs2_databuf_lops);
	bh->b_private = bd;

	if (meta)
		unlock_page(bh->b_page);
}

/**
 * gfs2_meta_wipe - make inode's buffers so they aren't dirty/pinned anymore
 * @ip: the inode who owns the buffers
 * @bstart: the first buffer in the run
 * @blen: the number of buffers in the run
 *
 */

void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct buffer_head *bh;

	while (blen) {
		bh = getbuf(ip->i_gl, bstart, NO_CREATE);
		if (bh) {
			struct gfs2_bufdata *bd = bh->b_private;

			if (test_clear_buffer_pinned(bh)) {
				struct gfs2_trans *tr = current->journal_info;
				struct gfs2_inode *bh_ip =
					GFS2_I(bh->b_page->mapping->host);

				gfs2_log_lock(sdp);
				list_del_init(&bd->bd_le.le_list);
				gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
				sdp->sd_log_num_buf--;
				gfs2_log_unlock(sdp);
				if (bh_ip->i_inode.i_private != NULL)
					tr->tr_num_databuf_rm++;
				else
					tr->tr_num_buf_rm++;
				brelse(bh);
			}
			if (bd) {
				gfs2_log_lock(sdp);
				if (bd->bd_ail) {
					u64 blkno = bh->b_blocknr;
					bd->bd_ail = NULL;
					list_del(&bd->bd_ail_st_list);
					list_del(&bd->bd_ail_gl_list);
					atomic_dec(&bd->bd_gl->gl_ail_count);
					brelse(bh);
					gfs2_log_unlock(sdp);
					gfs2_trans_add_revoke(sdp, blkno);
				} else
					gfs2_log_unlock(sdp);
			}

			lock_buffer(bh);
			clear_buffer_dirty(bh);
			clear_buffer_uptodate(bh);
			unlock_buffer(bh);

			brelse(bh);
		}

		bstart++;
		blen--;
	}
}

/**
 * gfs2_meta_cache_flush - get rid of any references on buffers for this inode
 * @ip: The GFS2 inode
 *
 * This releases buffers that are in the most-recently-used array of
 * blocks used for indirect block addressing for this inode.
 */

void gfs2_meta_cache_flush(struct gfs2_inode *ip)
{
	struct buffer_head **bh_slot;
	unsigned int x;

	spin_lock(&ip->i_spin);

	for (x = 0; x < GFS2_MAX_META_HEIGHT; x++) {
		bh_slot = &ip->i_cache[x];
		if (!*bh_slot)
			break;
		brelse(*bh_slot);
		*bh_slot = NULL;
	}

	spin_unlock(&ip->i_spin);
}

/**
 * gfs2_meta_indirect_buffer - Get a metadata buffer
 * @ip: The GFS2 inode
 * @height: The level of this buf in the metadata (indir addr) tree (if any)
 * @num: The block number (device relative) of the buffer
 * @new: Non-zero if we may create a new buffer
 * @bhp: the buffer is returned here
 *
 * Try to use the gfs2_inode's MRU metadata tree cache.
 *
 * Returns: errno
 */

int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
			      int new, struct buffer_head **bhp)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct gfs2_glock *gl = ip->i_gl;
	struct buffer_head *bh = NULL, **bh_slot = ip->i_cache + height;
	int in_cache = 0;

	BUG_ON(!gl);
	BUG_ON(!sdp);

	spin_lock(&ip->i_spin);
	if (*bh_slot && (*bh_slot)->b_blocknr == num) {
		bh = *bh_slot;
		get_bh(bh);
		in_cache = 1;
	}
	spin_unlock(&ip->i_spin);

	if (!bh)
		bh = getbuf(gl, num, CREATE);

	if (!bh)
		return -ENOBUFS;

	if (new) {
		if (gfs2_assert_warn(sdp, height))
			goto err;
		meta_prep_new(bh);
		gfs2_trans_add_bh(ip->i_gl, bh, 1);
		gfs2_metatype_set(bh, GFS2_METATYPE_IN, GFS2_FORMAT_IN);
		gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header));
	} else {
		u32 mtype = height ? GFS2_METATYPE_IN : GFS2_METATYPE_DI;
		if (!buffer_uptodate(bh)) {
			ll_rw_block(READ_META, 1, &bh);
			if (gfs2_meta_wait(sdp, bh))
				goto err;
		}
		if (gfs2_metatype_check(sdp, bh, mtype))
			goto err;
	}

	if (!in_cache) {
		spin_lock(&ip->i_spin);
		if (*bh_slot)
			brelse(*bh_slot);
		*bh_slot = bh;
		get_bh(bh);
		spin_unlock(&ip->i_spin);
	}

	*bhp = bh;
	return 0;
err:
	brelse(bh);
	return -EIO;
}

/**
 * gfs2_meta_ra - start readahead on an extent of a file
 * @gl: the glock the blocks belong to
 * @dblock: the starting disk block
 * @extlen: the number of blocks in the extent
 *
 * returns: the first buffer in the extent
 */

struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct buffer_head *first_bh, *bh;
	u32 max_ra = gfs2_tune_get(sdp, gt_max_readahead) >>
			  sdp->sd_sb.sb_bsize_shift;

	BUG_ON(!extlen);

	if (max_ra < 1)
		max_ra = 1;
	if (extlen > max_ra)
		extlen = max_ra;

	first_bh = getbuf(gl, dblock, CREATE);

	if (buffer_uptodate(first_bh))
		goto out;
	if (!buffer_locked(first_bh))
		ll_rw_block(READ_META, 1, &first_bh);

	dblock++;
	extlen--;

	while (extlen) {
		bh = getbuf(gl, dblock, CREATE);

		if (!buffer_uptodate(bh) && !buffer_locked(bh))
			ll_rw_block(READA, 1, &bh);
		brelse(bh);
		dblock++;
		extlen--;
		if (!buffer_locked(first_bh) && buffer_uptodate(first_bh))
			goto out;
	}

	wait_on_buffer(first_bh);
out:
	return first_bh;
}
Commit	Line	Data
b3b94faa DT	1	/*
b3b94faa DT	2	* Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
3a8a9a10	3	* Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
b3b94faa DT	4	*
	5	* This copyrighted material is made available to anyone wishing to use,
	6	* modify, copy, or redistribute it subject to the terms and conditions
e9fc2aa0	7	* of the GNU General Public License version 2.
b3b94faa DT	8	*/
	9
	10	#include <linux/sched.h>
	11	#include <linux/slab.h>
	12	#include <linux/spinlock.h>
	13	#include <linux/completion.h>
	14	#include <linux/buffer_head.h>
	15	#include <linux/mm.h>
	16	#include <linux/pagemap.h>
	17	#include <linux/writeback.h>
	18	#include <linux/swap.h>
	19	#include <linux/delay.h>
2e565bb6	20	#include <linux/bio.h>
5c676f6d	21	#include <linux/gfs2_ondisk.h>
7d308590	22	#include <linux/lm_interface.h>
b3b94faa DT	23
b3b94faa DT	24	#include "gfs2.h"
5c676f6d	25	#include "incore.h"
b3b94faa DT	26	#include "glock.h"
	27	#include "glops.h"
	28	#include "inode.h"
	29	#include "log.h"
	30	#include "lops.h"
	31	#include "meta_io.h"
	32	#include "rgrp.h"
	33	#include "trans.h"
5c676f6d	34	#include "util.h"
4340fe62	35	#include "ops_address.h"
b3b94faa	36
b3b94faa DT	37	static int aspace_get_block(struct inode *inode, sector_t lblock,
	38	struct buffer_head *bh_result, int create)
	39	{
5c676f6d	40	gfs2_assert_warn(inode->i_sb->s_fs_info, 0);
b3b94faa DT	41	return -EOPNOTSUPP;
	42	}
	43
	44	static int gfs2_aspace_writepage(struct page *page,
	45	struct writeback_control *wbc)
	46	{
	47	return block_write_full_page(page, aspace_get_block, wbc);
	48	}
	49
66de045d	50	static const struct address_space_operations aspace_aops = {
b3b94faa	51	.writepage = gfs2_aspace_writepage,
4340fe62	52	.releasepage = gfs2_releasepage,
b3b94faa DT	53	};
	54
	55	/**
	56	* gfs2_aspace_get - Create and initialize a struct inode structure
	57	* @sdp: the filesystem the aspace is in
	58	*
	59	* Right now a struct inode is just a struct inode. Maybe Linux
	60	* will supply a more lightweight address space construct (that works)
	61	* in the future.
	62	*
	63	* Make sure pages/buffers in this aspace aren't in high memory.
	64	*
	65	* Returns: the aspace
	66	*/
	67
	68	struct inode gfs2_aspace_get(struct gfs2_sbd sdp)
	69	{
	70	struct inode *aspace;
	71
	72	aspace = new_inode(sdp->sd_vfs);
	73	if (aspace) {
f3bba03f	74	mapping_set_gfp_mask(aspace->i_mapping, GFP_NOFS);
b3b94faa DT	75	aspace->i_mapping->a_ops = &aspace_aops;
b3b94faa DT	76	aspace->i_size = ~0ULL;
bba9dfd8	77	aspace->i_private = NULL;
b3b94faa DT	78	insert_inode_hash(aspace);
b3b94faa DT	79	}
b3b94faa DT	80	return aspace;
	81	}
	82
	83	void gfs2_aspace_put(struct inode *aspace)
	84	{
	85	remove_inode_hash(aspace);
	86	iput(aspace);
	87	}
	88
b3b94faa DT	89	/**
	90	* gfs2_meta_inval - Invalidate all buffers associated with a glock
	91	* @gl: the glock
	92	*
	93	*/
	94
	95	void gfs2_meta_inval(struct gfs2_glock *gl)
	96	{
	97	struct gfs2_sbd *sdp = gl->gl_sbd;
	98	struct inode *aspace = gl->gl_aspace;
	99	struct address_space *mapping = gl->gl_aspace->i_mapping;
	100
	101	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
	102
	103	atomic_inc(&aspace->i_writecount);
	104	truncate_inode_pages(mapping, 0);
	105	atomic_dec(&aspace->i_writecount);
	106
	107	gfs2_assert_withdraw(sdp, !mapping->nrpages);
	108	}
	109
	110	/**
	111	* gfs2_meta_sync - Sync all buffers associated with a glock
	112	* @gl: The glock
b3b94faa DT	113	*
	114	*/
	115
7276b3b0	116	void gfs2_meta_sync(struct gfs2_glock *gl)
b3b94faa DT	117	{
b3b94faa DT	118	struct address_space *mapping = gl->gl_aspace->i_mapping;
7276b3b0	119	int error;
b3b94faa	120
7276b3b0 SW	121	filemap_fdatawrite(mapping);
7276b3b0 SW	122	error = filemap_fdatawait(mapping);
b3b94faa DT	123
	124	if (error)
	125	gfs2_io_error(gl->gl_sbd);
	126	}
	127
	128	/**
	129	* getbuf - Get a buffer with a given address space
cb4c0313	130	* @gl: the glock
b3b94faa DT	131	* @blkno: the block number (filesystem scope)
	132	* @create: 1 if the buffer should be created
	133	*
	134	* Returns: the buffer
	135	*/
	136
cb4c0313	137	static struct buffer_head getbuf(struct gfs2_glock gl, u64 blkno, int create)
b3b94faa	138	{
cb4c0313 SW	139	struct address_space *mapping = gl->gl_aspace->i_mapping;
cb4c0313 SW	140	struct gfs2_sbd *sdp = gl->gl_sbd;
b3b94faa DT	141	struct page *page;
	142	struct buffer_head *bh;
	143	unsigned int shift;
	144	unsigned long index;
	145	unsigned int bufnum;
	146
	147	shift = PAGE_CACHE_SHIFT - sdp->sd_sb.sb_bsize_shift;
	148	index = blkno >> shift; /* convert block to page */
	149	bufnum = blkno - (index << shift); /* block buf index within page */
	150
	151	if (create) {
	152	for (;;) {
cb4c0313	153	page = grab_cache_page(mapping, index);
b3b94faa DT	154	if (page)
	155	break;
	156	yield();
	157	}
	158	} else {
cb4c0313	159	page = find_lock_page(mapping, index);
b3b94faa DT	160	if (!page)
	161	return NULL;
	162	}
	163
	164	if (!page_has_buffers(page))
	165	create_empty_buffers(page, sdp->sd_sb.sb_bsize, 0);
	166
	167	/* Locate header for our buffer within our page */
	168	for (bh = page_buffers(page); bufnum--; bh = bh->b_this_page)
	169	/* Do nothing */;
	170	get_bh(bh);
	171
	172	if (!buffer_mapped(bh))
	173	map_bh(bh, sdp->sd_vfs, blkno);
	174
	175	unlock_page(page);
	176	mark_page_accessed(page);
	177	page_cache_release(page);
	178
	179	return bh;
	180	}
	181
	182	static void meta_prep_new(struct buffer_head *bh)
	183	{
	184	struct gfs2_meta_header mh = (struct gfs2_meta_header )bh->b_data;
	185
	186	lock_buffer(bh);
	187	clear_buffer_dirty(bh);
	188	set_buffer_uptodate(bh);
	189	unlock_buffer(bh);
	190
	191	mh->mh_magic = cpu_to_be32(GFS2_MAGIC);
	192	}
	193
	194	/**
	195	* gfs2_meta_new - Get a block
	196	* @gl: The glock associated with this block
	197	* @blkno: The block number
	198	*
	199	* Returns: The buffer
	200	*/
	201
cd915493	202	struct buffer_head gfs2_meta_new(struct gfs2_glock gl, u64 blkno)
b3b94faa DT	203	{
b3b94faa DT	204	struct buffer_head *bh;
cb4c0313	205	bh = getbuf(gl, blkno, CREATE);
b3b94faa DT	206	meta_prep_new(bh);
	207	return bh;
	208	}
	209
	210	/**
	211	* gfs2_meta_read - Read a block from disk
	212	* @gl: The glock covering the block
	213	* @blkno: The block number
7276b3b0	214	* @flags: flags
b3b94faa DT	215	* @bhp: the place where the buffer is returned (NULL on failure)
	216	*
	217	* Returns: errno
	218	*/
	219
cd915493	220	int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags,
b3b94faa DT	221	struct buffer_head **bhp)
b3b94faa DT	222	{
cb4c0313	223	*bhp = getbuf(gl, blkno, CREATE);
7276b3b0	224	if (!buffer_uptodate(*bhp))
2e565bb6	225	ll_rw_block(READ_META, 1, bhp);
7276b3b0 SW	226	if (flags & DIO_WAIT) {
	227	int error = gfs2_meta_wait(gl->gl_sbd, *bhp);
	228	if (error) {
	229	brelse(*bhp);
	230	return error;
	231	}
	232	}
b3b94faa	233
7276b3b0	234	return 0;
b3b94faa DT	235	}
	236
	237	/**
7276b3b0	238	* gfs2_meta_wait - Reread a block from disk
b3b94faa	239	* @sdp: the filesystem
7276b3b0	240	* @bh: The block to wait for
b3b94faa DT	241	*
	242	* Returns: errno
	243	*/
	244
7276b3b0	245	int gfs2_meta_wait(struct gfs2_sbd sdp, struct buffer_head bh)
b3b94faa DT	246	{
	247	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
	248	return -EIO;
	249
7276b3b0	250	wait_on_buffer(bh);
b3b94faa	251
7276b3b0 SW	252	if (!buffer_uptodate(bh)) {
	253	struct gfs2_trans *tr = current->journal_info;
	254	if (tr && tr->tr_touched)
	255	gfs2_io_error_bh(sdp, bh);
	256	return -EIO;
b3b94faa	257	}
7276b3b0 SW	258	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
7276b3b0 SW	259	return -EIO;
b3b94faa DT	260
	261	return 0;
	262	}
	263
	264	/**
586dfdaa	265	* gfs2_attach_bufdata - attach a struct gfs2_bufdata structure to a buffer
b3b94faa DT	266	* @gl: the glock the buffer belongs to
b3b94faa DT	267	* @bh: The buffer to be attached to
586dfdaa	268	* @meta: Flag to indicate whether its metadata or not
b3b94faa DT	269	*/
b3b94faa DT	270
568f4c96 SW	271	void gfs2_attach_bufdata(struct gfs2_glock gl, struct buffer_head bh,
568f4c96 SW	272	int meta)
b3b94faa DT	273	{
	274	struct gfs2_bufdata *bd;
	275
18ec7d5c SW	276	if (meta)
18ec7d5c SW	277	lock_page(bh->b_page);
b3b94faa	278
5c676f6d	279	if (bh->b_private) {
18ec7d5c SW	280	if (meta)
18ec7d5c SW	281	unlock_page(bh->b_page);
b3b94faa DT	282	return;
	283	}
	284
c3762229	285	bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS \| __GFP_NOFAIL),
b3b94faa DT	286	bd->bd_bh = bh;
	287	bd->bd_gl = gl;
	288
	289	INIT_LIST_HEAD(&bd->bd_list_tr);
82ffa516	290	if (meta)
586dfdaa	291	lops_init_le(&bd->bd_le, &gfs2_buf_lops);
82ffa516	292	else
586dfdaa	293	lops_init_le(&bd->bd_le, &gfs2_databuf_lops);
5c676f6d	294	bh->b_private = bd;
b3b94faa	295
18ec7d5c SW	296	if (meta)
18ec7d5c SW	297	unlock_page(bh->b_page);
b3b94faa DT	298	}
b3b94faa DT	299
b3b94faa DT	300	/**
	301	* gfs2_meta_wipe - make inode's buffers so they aren't dirty/pinned anymore
	302	* @ip: the inode who owns the buffers
	303	* @bstart: the first buffer in the run
	304	* @blen: the number of buffers in the run
	305	*
	306	*/
	307
cd915493	308	void gfs2_meta_wipe(struct gfs2_inode *ip, u64 bstart, u32 blen)
b3b94faa	309	{
feaa7bba	310	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
b3b94faa DT	311	struct buffer_head *bh;
	312
	313	while (blen) {
cb4c0313	314	bh = getbuf(ip->i_gl, bstart, NO_CREATE);
b3b94faa	315	if (bh) {
5c676f6d	316	struct gfs2_bufdata *bd = bh->b_private;
b3b94faa DT	317
b3b94faa DT	318	if (test_clear_buffer_pinned(bh)) {
5c676f6d	319	struct gfs2_trans *tr = current->journal_info;
2332c443 RP	320	struct gfs2_inode *bh_ip =
	321	GFS2_I(bh->b_page->mapping->host);
	322
b3b94faa DT	323	gfs2_log_lock(sdp);
	324	list_del_init(&bd->bd_le.le_list);
	325	gfs2_assert_warn(sdp, sdp->sd_log_num_buf);
	326	sdp->sd_log_num_buf--;
	327	gfs2_log_unlock(sdp);
2332c443 RP	328	if (bh_ip->i_inode.i_private != NULL)
	329	tr->tr_num_databuf_rm++;
	330	else
	331	tr->tr_num_buf_rm++;
b3b94faa DT	332	brelse(bh);
	333	}
	334	if (bd) {
	335	gfs2_log_lock(sdp);
	336	if (bd->bd_ail) {
cd915493	337	u64 blkno = bh->b_blocknr;
b3b94faa DT	338	bd->bd_ail = NULL;
	339	list_del(&bd->bd_ail_st_list);
	340	list_del(&bd->bd_ail_gl_list);
	341	atomic_dec(&bd->bd_gl->gl_ail_count);
	342	brelse(bh);
	343	gfs2_log_unlock(sdp);
	344	gfs2_trans_add_revoke(sdp, blkno);
	345	} else
	346	gfs2_log_unlock(sdp);
	347	}
	348
	349	lock_buffer(bh);
	350	clear_buffer_dirty(bh);
	351	clear_buffer_uptodate(bh);
	352	unlock_buffer(bh);
	353
	354	brelse(bh);
	355	}
	356
	357	bstart++;
	358	blen--;
	359	}
	360	}
	361
	362	/**
	363	* gfs2_meta_cache_flush - get rid of any references on buffers for this inode
	364	* @ip: The GFS2 inode
	365	*
	366	* This releases buffers that are in the most-recently-used array of
	367	* blocks used for indirect block addressing for this inode.
	368	*/
	369
	370	void gfs2_meta_cache_flush(struct gfs2_inode *ip)
	371	{
	372	struct buffer_head **bh_slot;
	373	unsigned int x;
	374
	375	spin_lock(&ip->i_spin);
	376
	377	for (x = 0; x < GFS2_MAX_META_HEIGHT; x++) {
	378	bh_slot = &ip->i_cache[x];
	379	if (!*bh_slot)
	380	break;
	381	brelse(*bh_slot);
	382	*bh_slot = NULL;
	383	}
	384
	385	spin_unlock(&ip->i_spin);
	386	}
	387
	388	/**
	389	* gfs2_meta_indirect_buffer - Get a metadata buffer
	390	* @ip: The GFS2 inode
	391	* @height: The level of this buf in the metadata (indir addr) tree (if any)
	392	* @num: The block number (device relative) of the buffer
	393	* @new: Non-zero if we may create a new buffer
	394	* @bhp: the buffer is returned here
	395	*
	396	* Try to use the gfs2_inode's MRU metadata tree cache.
	397	*
	398	* Returns: errno
	399	*/
	400
cd915493	401	int gfs2_meta_indirect_buffer(struct gfs2_inode *ip, int height, u64 num,
b3b94faa DT	402	int new, struct buffer_head **bhp)
b3b94faa DT	403	{
7276b3b0 SW	404	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	405	struct gfs2_glock *gl = ip->i_gl;
	406	struct buffer_head bh = NULL, *bh_slot = ip->i_cache + height;
	407	int in_cache = 0;
b3b94faa	408
b004157a SW	409	BUG_ON(!gl);
	410	BUG_ON(!sdp);
	411
b3b94faa	412	spin_lock(&ip->i_spin);
7276b3b0 SW	413	if (bh_slot && (bh_slot)->b_blocknr == num) {
	414	bh = *bh_slot;
	415	get_bh(bh);
	416	in_cache = 1;
b3b94faa DT	417	}
	418	spin_unlock(&ip->i_spin);
	419
7276b3b0	420	if (!bh)
cb4c0313	421	bh = getbuf(gl, num, CREATE);
b3b94faa	422
7276b3b0 SW	423	if (!bh)
7276b3b0 SW	424	return -ENOBUFS;
b3b94faa DT	425
b3b94faa DT	426	if (new) {
7276b3b0 SW	427	if (gfs2_assert_warn(sdp, height))
	428	goto err;
	429	meta_prep_new(bh);
d4e9c4c3	430	gfs2_trans_add_bh(ip->i_gl, bh, 1);
b3b94faa DT	431	gfs2_metatype_set(bh, GFS2_METATYPE_IN, GFS2_FORMAT_IN);
b3b94faa DT	432	gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header));
7276b3b0 SW	433	} else {
	434	u32 mtype = height ? GFS2_METATYPE_IN : GFS2_METATYPE_DI;
	435	if (!buffer_uptodate(bh)) {
2e565bb6	436	ll_rw_block(READ_META, 1, &bh);
7276b3b0 SW	437	if (gfs2_meta_wait(sdp, bh))
	438	goto err;
	439	}
	440	if (gfs2_metatype_check(sdp, bh, mtype))
	441	goto err;
	442	}
b3b94faa	443
7276b3b0 SW	444	if (!in_cache) {
	445	spin_lock(&ip->i_spin);
	446	if (*bh_slot)
	447	brelse(*bh_slot);
	448	*bh_slot = bh;
	449	get_bh(bh);
	450	spin_unlock(&ip->i_spin);
b3b94faa DT	451	}
	452
	453	*bhp = bh;
b3b94faa	454	return 0;
7276b3b0 SW	455	err:
	456	brelse(bh);
	457	return -EIO;
b3b94faa DT	458	}
	459
	460	/**
	461	* gfs2_meta_ra - start readahead on an extent of a file
	462	* @gl: the glock the blocks belong to
	463	* @dblock: the starting disk block
	464	* @extlen: the number of blocks in the extent
	465	*
7276b3b0	466	* returns: the first buffer in the extent
b3b94faa DT	467	*/
b3b94faa DT	468
7276b3b0	469	struct buffer_head gfs2_meta_ra(struct gfs2_glock gl, u64 dblock, u32 extlen)
b3b94faa DT	470	{
b3b94faa DT	471	struct gfs2_sbd *sdp = gl->gl_sbd;
b3b94faa	472	struct buffer_head first_bh, bh;
cd915493	473	u32 max_ra = gfs2_tune_get(sdp, gt_max_readahead) >>
568f4c96	474	sdp->sd_sb.sb_bsize_shift;
b3b94faa	475
7276b3b0 SW	476	BUG_ON(!extlen);
	477
	478	if (max_ra < 1)
	479	max_ra = 1;
b3b94faa DT	480	if (extlen > max_ra)
	481	extlen = max_ra;
	482
cb4c0313	483	first_bh = getbuf(gl, dblock, CREATE);
b3b94faa DT	484
	485	if (buffer_uptodate(first_bh))
	486	goto out;
7276b3b0	487	if (!buffer_locked(first_bh))
2e565bb6	488	ll_rw_block(READ_META, 1, &first_bh);
b3b94faa DT	489
	490	dblock++;
	491	extlen--;
	492
	493	while (extlen) {
cb4c0313	494	bh = getbuf(gl, dblock, CREATE);
b3b94faa	495
7276b3b0 SW	496	if (!buffer_uptodate(bh) && !buffer_locked(bh))
	497	ll_rw_block(READA, 1, &bh);
	498	brelse(bh);
b3b94faa DT	499	dblock++;
b3b94faa DT	500	extlen--;
7276b3b0 SW	501	if (!buffer_locked(first_bh) && buffer_uptodate(first_bh))
7276b3b0 SW	502	goto out;
b3b94faa DT	503	}
b3b94faa DT	504
7276b3b0	505	wait_on_buffer(first_bh);
a91ea69f	506	out:
7276b3b0	507	return first_bh;
b3b94faa DT	508	}
b3b94faa DT	509