[linux.git] / fs / gfs2 / glops.c

/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
 * Copyright (C) 2004-2008 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/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
#include <linux/gfs2_ondisk.h>
#include <linux/bio.h>
#include <linux/posix_acl.h>

#include "gfs2.h"
#include "incore.h"
#include "bmap.h"
#include "glock.h"
#include "glops.h"
#include "inode.h"
#include "log.h"
#include "meta_io.h"
#include "recovery.h"
#include "rgrp.h"
#include "util.h"
#include "trans.h"

/**
 * __gfs2_ail_flush - remove all buffers for a given lock from the AIL
 * @gl: the glock
 *
 * None of the buffers should be dirty, locked, or pinned.
 */

static void __gfs2_ail_flush(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct list_head *head = &gl->gl_ail_list;
	struct gfs2_bufdata *bd;
	struct buffer_head *bh;

	spin_lock(&sdp->sd_ail_lock);
	while (!list_empty(head)) {
		bd = list_entry(head->next, struct gfs2_bufdata,
				bd_ail_gl_list);
		bh = bd->bd_bh;
		gfs2_remove_from_ail(bd);
		bd->bd_bh = NULL;
		bh->b_private = NULL;
		spin_unlock(&sdp->sd_ail_lock);

		bd->bd_blkno = bh->b_blocknr;
		gfs2_log_lock(sdp);
		gfs2_assert_withdraw(sdp, !buffer_busy(bh));
		gfs2_trans_add_revoke(sdp, bd);
		gfs2_log_unlock(sdp);

		spin_lock(&sdp->sd_ail_lock);
	}
	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
	spin_unlock(&sdp->sd_ail_lock);
}


static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct gfs2_trans tr;

	memset(&tr, 0, sizeof(tr));
	tr.tr_revokes = atomic_read(&gl->gl_ail_count);

	if (!tr.tr_revokes)
		return;

	/* A shortened, inline version of gfs2_trans_begin() */
	tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes, sizeof(u64));
	tr.tr_ip = (unsigned long)__builtin_return_address(0);
	INIT_LIST_HEAD(&tr.tr_list_buf);
	gfs2_log_reserve(sdp, tr.tr_reserved);
	BUG_ON(current->journal_info);
	current->journal_info = &tr;

	__gfs2_ail_flush(gl);

	gfs2_trans_end(sdp);
	gfs2_log_flush(sdp, NULL);
}

void gfs2_ail_flush(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	unsigned int revokes = atomic_read(&gl->gl_ail_count);
	int ret;

	if (!revokes)
		return;

	ret = gfs2_trans_begin(sdp, 0, revokes);
	if (ret)
		return;
	__gfs2_ail_flush(gl);
	gfs2_trans_end(sdp);
	gfs2_log_flush(sdp, NULL);
}

/**
 * rgrp_go_sync - sync out the metadata for this glock
 * @gl: the glock
 *
 * Called when demoting or unlocking an EX glock.  We must flush
 * to disk all dirty buffers/pages relating to this glock, and must not
 * not return to caller to demote/unlock the glock until I/O is complete.
 */

static void rgrp_go_sync(struct gfs2_glock *gl)
{
	struct address_space *metamapping = gfs2_glock2aspace(gl);
	int error;

	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
		return;
	BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE);

	gfs2_log_flush(gl->gl_sbd, gl);
	filemap_fdatawrite(metamapping);
	error = filemap_fdatawait(metamapping);
        mapping_set_error(metamapping, error);
	gfs2_ail_empty_gl(gl);
}

/**
 * rgrp_go_inval - invalidate the metadata for this glock
 * @gl: the glock
 * @flags:
 *
 * We never used LM_ST_DEFERRED with resource groups, so that we
 * should always see the metadata flag set here.
 *
 */

static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
{
	struct address_space *mapping = gfs2_glock2aspace(gl);

	BUG_ON(!(flags & DIO_METADATA));
	gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count));
	truncate_inode_pages(mapping, 0);

	if (gl->gl_object) {
		struct gfs2_rgrpd *rgd = (struct gfs2_rgrpd *)gl->gl_object;
		rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
	}
}

/**
 * inode_go_sync - Sync the dirty data and/or metadata for an inode glock
 * @gl: the glock protecting the inode
 *
 */

static void inode_go_sync(struct gfs2_glock *gl)
{
	struct gfs2_inode *ip = gl->gl_object;
	struct address_space *metamapping = gfs2_glock2aspace(gl);
	int error;

	if (ip && !S_ISREG(ip->i_inode.i_mode))
		ip = NULL;
	if (ip && test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags))
		unmap_shared_mapping_range(ip->i_inode.i_mapping, 0, 0);
	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
		return;

	BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE);

	gfs2_log_flush(gl->gl_sbd, gl);
	filemap_fdatawrite(metamapping);
	if (ip) {
		struct address_space *mapping = ip->i_inode.i_mapping;
		filemap_fdatawrite(mapping);
		error = filemap_fdatawait(mapping);
		mapping_set_error(mapping, error);
	}
	error = filemap_fdatawait(metamapping);
	mapping_set_error(metamapping, error);
	gfs2_ail_empty_gl(gl);
	/*
	 * Writeback of the data mapping may cause the dirty flag to be set
	 * so we have to clear it again here.
	 */
	smp_mb__before_clear_bit();
	clear_bit(GLF_DIRTY, &gl->gl_flags);
}

/**
 * inode_go_inval - prepare a inode glock to be released
 * @gl: the glock
 * @flags:
 * 
 * Normally we invlidate everything, but if we are moving into
 * LM_ST_DEFERRED from LM_ST_SHARED or LM_ST_EXCLUSIVE then we
 * can keep hold of the metadata, since it won't have changed.
 *
 */

static void inode_go_inval(struct gfs2_glock *gl, int flags)
{
	struct gfs2_inode *ip = gl->gl_object;

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

	if (flags & DIO_METADATA) {
		struct address_space *mapping = gfs2_glock2aspace(gl);
		truncate_inode_pages(mapping, 0);
		if (ip) {
			set_bit(GIF_INVALID, &ip->i_flags);
			forget_all_cached_acls(&ip->i_inode);
		}
	}

	if (ip == GFS2_I(gl->gl_sbd->sd_rindex)) {
		gfs2_log_flush(gl->gl_sbd, NULL);
		gl->gl_sbd->sd_rindex_uptodate = 0;
	}
	if (ip && S_ISREG(ip->i_inode.i_mode))
		truncate_inode_pages(ip->i_inode.i_mapping, 0);
}

/**
 * inode_go_demote_ok - Check to see if it's ok to unlock an inode glock
 * @gl: the glock
 *
 * Returns: 1 if it's ok
 */

static int inode_go_demote_ok(const struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct gfs2_holder *gh;

	if (sdp->sd_jindex == gl->gl_object || sdp->sd_rindex == gl->gl_object)
		return 0;

	if (!list_empty(&gl->gl_holders)) {
		gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list);
		if (gh->gh_list.next != &gl->gl_holders)
			return 0;
	}

	return 1;
}

/**
 * gfs2_set_nlink - Set the inode's link count based on on-disk info
 * @inode: The inode in question
 * @nlink: The link count
 *
 * If the link count has hit zero, it must never be raised, whatever the
 * on-disk inode might say. When new struct inodes are created the link
 * count is set to 1, so that we can safely use this test even when reading
 * in on disk information for the first time.
 */

static void gfs2_set_nlink(struct inode *inode, u32 nlink)
{
	/*
	 * We will need to review setting the nlink count here in the
	 * light of the forthcoming ro bind mount work. This is a reminder
	 * to do that.
	 */
	if ((inode->i_nlink != nlink) && (inode->i_nlink != 0)) {
		if (nlink == 0)
			clear_nlink(inode);
		else
			inode->i_nlink = nlink;
	}
}

static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
{
	const struct gfs2_dinode *str = buf;
	struct timespec atime;
	u16 height, depth;

	if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr)))
		goto corrupt;
	ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino);
	ip->i_inode.i_mode = be32_to_cpu(str->di_mode);
	ip->i_inode.i_rdev = 0;
	switch (ip->i_inode.i_mode & S_IFMT) {
	case S_IFBLK:
	case S_IFCHR:
		ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major),
					   be32_to_cpu(str->di_minor));
		break;
	};

	ip->i_inode.i_uid = be32_to_cpu(str->di_uid);
	ip->i_inode.i_gid = be32_to_cpu(str->di_gid);
	gfs2_set_nlink(&ip->i_inode, be32_to_cpu(str->di_nlink));
	i_size_write(&ip->i_inode, be64_to_cpu(str->di_size));
	gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks));
	atime.tv_sec = be64_to_cpu(str->di_atime);
	atime.tv_nsec = be32_to_cpu(str->di_atime_nsec);
	if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0)
		ip->i_inode.i_atime = atime;
	ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime);
	ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec);
	ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime);
	ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec);

	ip->i_goal = be64_to_cpu(str->di_goal_meta);
	ip->i_generation = be64_to_cpu(str->di_generation);

	ip->i_diskflags = be32_to_cpu(str->di_flags);
	gfs2_set_inode_flags(&ip->i_inode);
	height = be16_to_cpu(str->di_height);
	if (unlikely(height > GFS2_MAX_META_HEIGHT))
		goto corrupt;
	ip->i_height = (u8)height;

	depth = be16_to_cpu(str->di_depth);
	if (unlikely(depth > GFS2_DIR_MAX_DEPTH))
		goto corrupt;
	ip->i_depth = (u8)depth;
	ip->i_entries = be32_to_cpu(str->di_entries);

	ip->i_eattr = be64_to_cpu(str->di_eattr);
	if (S_ISREG(ip->i_inode.i_mode))
		gfs2_set_aops(&ip->i_inode);

	return 0;
corrupt:
	gfs2_consist_inode(ip);
	return -EIO;
}

/**
 * gfs2_inode_refresh - Refresh the incore copy of the dinode
 * @ip: The GFS2 inode
 *
 * Returns: errno
 */

int gfs2_inode_refresh(struct gfs2_inode *ip)
{
	struct buffer_head *dibh;
	int error;

	error = gfs2_meta_inode_buffer(ip, &dibh);
	if (error)
		return error;

	if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), dibh, GFS2_METATYPE_DI)) {
		brelse(dibh);
		return -EIO;
	}

	error = gfs2_dinode_in(ip, dibh->b_data);
	brelse(dibh);
	clear_bit(GIF_INVALID, &ip->i_flags);

	return error;
}

/**
 * inode_go_lock - operation done after an inode lock is locked by a process
 * @gl: the glock
 * @flags:
 *
 * Returns: errno
 */

static int inode_go_lock(struct gfs2_holder *gh)
{
	struct gfs2_glock *gl = gh->gh_gl;
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct gfs2_inode *ip = gl->gl_object;
	int error = 0;

	if (!ip || (gh->gh_flags & GL_SKIP))
		return 0;

	if (test_bit(GIF_INVALID, &ip->i_flags)) {
		error = gfs2_inode_refresh(ip);
		if (error)
			return error;
	}

	if ((ip->i_diskflags & GFS2_DIF_TRUNC_IN_PROG) &&
	    (gl->gl_state == LM_ST_EXCLUSIVE) &&
	    (gh->gh_state == LM_ST_EXCLUSIVE)) {
		spin_lock(&sdp->sd_trunc_lock);
		if (list_empty(&ip->i_trunc_list))
			list_add(&sdp->sd_trunc_list, &ip->i_trunc_list);
		spin_unlock(&sdp->sd_trunc_lock);
		wake_up(&sdp->sd_quota_wait);
		return 1;
	}

	return error;
}

/**
 * inode_go_dump - print information about an inode
 * @seq: The iterator
 * @ip: the inode
 *
 * Returns: 0 on success, -ENOBUFS when we run out of space
 */

static int inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl)
{
	const struct gfs2_inode *ip = gl->gl_object;
	if (ip == NULL)
		return 0;
	gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu\n",
		  (unsigned long long)ip->i_no_formal_ino,
		  (unsigned long long)ip->i_no_addr,
		  IF2DT(ip->i_inode.i_mode), ip->i_flags,
		  (unsigned int)ip->i_diskflags,
		  (unsigned long long)i_size_read(&ip->i_inode));
	return 0;
}

/**
 * rgrp_go_lock - operation done after an rgrp lock is locked by
 *    a first holder on this node.
 * @gl: the glock
 * @flags:
 *
 * Returns: errno
 */

static int rgrp_go_lock(struct gfs2_holder *gh)
{
	return gfs2_rgrp_bh_get(gh->gh_gl->gl_object);
}

/**
 * rgrp_go_unlock - operation done before an rgrp lock is unlocked by
 *    a last holder on this node.
 * @gl: the glock
 * @flags:
 *
 */

static void rgrp_go_unlock(struct gfs2_holder *gh)
{
	gfs2_rgrp_bh_put(gh->gh_gl->gl_object);
}

/**
 * trans_go_sync - promote/demote the transaction glock
 * @gl: the glock
 * @state: the requested state
 * @flags:
 *
 */

static void trans_go_sync(struct gfs2_glock *gl)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;

	if (gl->gl_state != LM_ST_UNLOCKED &&
	    test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
		gfs2_meta_syncfs(sdp);
		gfs2_log_shutdown(sdp);
	}
}

/**
 * trans_go_xmote_bh - After promoting/demoting the transaction glock
 * @gl: the glock
 *
 */

static int trans_go_xmote_bh(struct gfs2_glock *gl, struct gfs2_holder *gh)
{
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode);
	struct gfs2_glock *j_gl = ip->i_gl;
	struct gfs2_log_header_host head;
	int error;

	if (test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
		j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);

		error = gfs2_find_jhead(sdp->sd_jdesc, &head);
		if (error)
			gfs2_consist(sdp);
		if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT))
			gfs2_consist(sdp);

		/*  Initialize some head of the log stuff  */
		if (!test_bit(SDF_SHUTDOWN, &sdp->sd_flags)) {
			sdp->sd_log_sequence = head.lh_sequence + 1;
			gfs2_log_pointers_init(sdp, head.lh_blkno);
		}
	}
	return 0;
}

/**
 * trans_go_demote_ok
 * @gl: the glock
 *
 * Always returns 0
 */

static int trans_go_demote_ok(const struct gfs2_glock *gl)
{
	return 0;
}

/**
 * iopen_go_callback - schedule the dcache entry for the inode to be deleted
 * @gl: the glock
 *
 * gl_spin lock is held while calling this
 */
static void iopen_go_callback(struct gfs2_glock *gl)
{
	struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object;
	struct gfs2_sbd *sdp = gl->gl_sbd;

	if (sdp->sd_vfs->s_flags & MS_RDONLY)
		return;

	if (gl->gl_demote_state == LM_ST_UNLOCKED &&
	    gl->gl_state == LM_ST_SHARED && ip) {
		gfs2_glock_hold(gl);
		if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
			gfs2_glock_put_nolock(gl);
	}
}

const struct gfs2_glock_operations gfs2_meta_glops = {
	.go_type = LM_TYPE_META,
};

const struct gfs2_glock_operations gfs2_inode_glops = {
	.go_xmote_th = inode_go_sync,
	.go_inval = inode_go_inval,
	.go_demote_ok = inode_go_demote_ok,
	.go_lock = inode_go_lock,
	.go_dump = inode_go_dump,
	.go_type = LM_TYPE_INODE,
	.go_min_hold_time = HZ / 5,
	.go_flags = GLOF_ASPACE,
};

const struct gfs2_glock_operations gfs2_rgrp_glops = {
	.go_xmote_th = rgrp_go_sync,
	.go_inval = rgrp_go_inval,
	.go_lock = rgrp_go_lock,
	.go_unlock = rgrp_go_unlock,
	.go_dump = gfs2_rgrp_dump,
	.go_type = LM_TYPE_RGRP,
	.go_min_hold_time = HZ / 5,
	.go_flags = GLOF_ASPACE,
};

const struct gfs2_glock_operations gfs2_trans_glops = {
	.go_xmote_th = trans_go_sync,
	.go_xmote_bh = trans_go_xmote_bh,
	.go_demote_ok = trans_go_demote_ok,
	.go_type = LM_TYPE_NONDISK,
};

const struct gfs2_glock_operations gfs2_iopen_glops = {
	.go_type = LM_TYPE_IOPEN,
	.go_callback = iopen_go_callback,
};

const struct gfs2_glock_operations gfs2_flock_glops = {
	.go_type = LM_TYPE_FLOCK,
};

const struct gfs2_glock_operations gfs2_nondisk_glops = {
	.go_type = LM_TYPE_NONDISK,
};

const struct gfs2_glock_operations gfs2_quota_glops = {
	.go_type = LM_TYPE_QUOTA,
};

const struct gfs2_glock_operations gfs2_journal_glops = {
	.go_type = LM_TYPE_JOURNAL,
};

const struct gfs2_glock_operations *gfs2_glops_list[] = {
	[LM_TYPE_META] = &gfs2_meta_glops,
	[LM_TYPE_INODE] = &gfs2_inode_glops,
	[LM_TYPE_RGRP] = &gfs2_rgrp_glops,
	[LM_TYPE_IOPEN] = &gfs2_iopen_glops,
	[LM_TYPE_FLOCK] = &gfs2_flock_glops,
	[LM_TYPE_NONDISK] = &gfs2_nondisk_glops,
	[LM_TYPE_QUOTA] = &gfs2_quota_glops,
	[LM_TYPE_JOURNAL] = &gfs2_journal_glops,
};
Commit	Line	Data
b3b94faa DT	1	/*
b3b94faa DT	2	* Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
cf45b752	3	* Copyright (C) 2004-2008 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	*/
b3b94faa DT	9
b3b94faa DT	10	#include <linux/spinlock.h>
	11	#include <linux/completion.h>
	12	#include <linux/buffer_head.h>
5c676f6d	13	#include <linux/gfs2_ondisk.h>
6802e340	14	#include <linux/bio.h>
c65f7fb5	15	#include <linux/posix_acl.h>
b3b94faa DT	16
b3b94faa DT	17	#include "gfs2.h"
5c676f6d	18	#include "incore.h"
b3b94faa DT	19	#include "bmap.h"
	20	#include "glock.h"
	21	#include "glops.h"
	22	#include "inode.h"
	23	#include "log.h"
	24	#include "meta_io.h"
b3b94faa DT	25	#include "recovery.h"
b3b94faa DT	26	#include "rgrp.h"
5c676f6d	27	#include "util.h"
ddacfaf7	28	#include "trans.h"
b3b94faa	29
ddacfaf7	30	/**
dba898b0	31	* __gfs2_ail_flush - remove all buffers for a given lock from the AIL
ddacfaf7 SW	32	* @gl: the glock
	33	*
	34	* None of the buffers should be dirty, locked, or pinned.
	35	*/
	36
dba898b0	37	static void __gfs2_ail_flush(struct gfs2_glock *gl)
ddacfaf7 SW	38	{
ddacfaf7 SW	39	struct gfs2_sbd *sdp = gl->gl_sbd;
ddacfaf7 SW	40	struct list_head *head = &gl->gl_ail_list;
	41	struct gfs2_bufdata *bd;
	42	struct buffer_head *bh;
d8348de0	43
d6a079e8	44	spin_lock(&sdp->sd_ail_lock);
ddacfaf7 SW	45	while (!list_empty(head)) {
	46	bd = list_entry(head->next, struct gfs2_bufdata,
	47	bd_ail_gl_list);
	48	bh = bd->bd_bh;
f91a0d3e	49	gfs2_remove_from_ail(bd);
1ad38c43 SW	50	bd->bd_bh = NULL;
1ad38c43 SW	51	bh->b_private = NULL;
380f7c65 SW	52	spin_unlock(&sdp->sd_ail_lock);
380f7c65 SW	53
1ad38c43	54	bd->bd_blkno = bh->b_blocknr;
d6a079e8	55	gfs2_log_lock(sdp);
1ad38c43 SW	56	gfs2_assert_withdraw(sdp, !buffer_busy(bh));
1ad38c43 SW	57	gfs2_trans_add_revoke(sdp, bd);
d6a079e8 DC	58	gfs2_log_unlock(sdp);
	59
	60	spin_lock(&sdp->sd_ail_lock);
ddacfaf7 SW	61	}
ddacfaf7 SW	62	gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count));
d6a079e8	63	spin_unlock(&sdp->sd_ail_lock);
dba898b0 SW	64	}
	65
	66
	67	static void gfs2_ail_empty_gl(struct gfs2_glock *gl)
	68	{
	69	struct gfs2_sbd *sdp = gl->gl_sbd;
	70	struct gfs2_trans tr;
	71
	72	memset(&tr, 0, sizeof(tr));
	73	tr.tr_revokes = atomic_read(&gl->gl_ail_count);
	74
	75	if (!tr.tr_revokes)
	76	return;
	77
	78	/* A shortened, inline version of gfs2_trans_begin() */
	79	tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes, sizeof(u64));
	80	tr.tr_ip = (unsigned long)__builtin_return_address(0);
	81	INIT_LIST_HEAD(&tr.tr_list_buf);
	82	gfs2_log_reserve(sdp, tr.tr_reserved);
	83	BUG_ON(current->journal_info);
	84	current->journal_info = &tr;
	85
	86	__gfs2_ail_flush(gl);
	87
	88	gfs2_trans_end(sdp);
	89	gfs2_log_flush(sdp, NULL);
	90	}
ddacfaf7	91
dba898b0 SW	92	void gfs2_ail_flush(struct gfs2_glock *gl)
	93	{
	94	struct gfs2_sbd *sdp = gl->gl_sbd;
	95	unsigned int revokes = atomic_read(&gl->gl_ail_count);
	96	int ret;
	97
	98	if (!revokes)
	99	return;
	100
	101	ret = gfs2_trans_begin(sdp, 0, revokes);
	102	if (ret)
	103	return;
	104	__gfs2_ail_flush(gl);
ddacfaf7 SW	105	gfs2_trans_end(sdp);
	106	gfs2_log_flush(sdp, NULL);
	107	}
ba7f7290 SW	108
ba7f7290 SW	109	/**
6bac243f	110	* rgrp_go_sync - sync out the metadata for this glock
b3b94faa	111	* @gl: the glock
b3b94faa DT	112	*
	113	* Called when demoting or unlocking an EX glock. We must flush
	114	* to disk all dirty buffers/pages relating to this glock, and must not
	115	* not return to caller to demote/unlock the glock until I/O is complete.
	116	*/
	117
6bac243f	118	static void rgrp_go_sync(struct gfs2_glock *gl)
b3b94faa	119	{
009d8518	120	struct address_space *metamapping = gfs2_glock2aspace(gl);
6bac243f SW	121	int error;
	122
	123	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
b5d32bea	124	return;
6bac243f	125	BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE);
b5d32bea	126
6bac243f SW	127	gfs2_log_flush(gl->gl_sbd, gl);
	128	filemap_fdatawrite(metamapping);
	129	error = filemap_fdatawait(metamapping);
	130	mapping_set_error(metamapping, error);
	131	gfs2_ail_empty_gl(gl);
b3b94faa DT	132	}
	133
	134	/**
6bac243f	135	* rgrp_go_inval - invalidate the metadata for this glock
b3b94faa DT	136	* @gl: the glock
	137	* @flags:
	138	*
6bac243f SW	139	* We never used LM_ST_DEFERRED with resource groups, so that we
	140	* should always see the metadata flag set here.
	141	*
b3b94faa DT	142	*/
b3b94faa DT	143
6bac243f	144	static void rgrp_go_inval(struct gfs2_glock *gl, int flags)
b3b94faa	145	{
009d8518	146	struct address_space *mapping = gfs2_glock2aspace(gl);
b3b94faa	147
6bac243f SW	148	BUG_ON(!(flags & DIO_METADATA));
	149	gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count));
	150	truncate_inode_pages(mapping, 0);
cf45b752	151
6bac243f SW	152	if (gl->gl_object) {
6bac243f SW	153	struct gfs2_rgrpd rgd = (struct gfs2_rgrpd )gl->gl_object;
cf45b752 BP	154	rgd->rd_flags &= ~GFS2_RDF_UPTODATE;
cf45b752 BP	155	}
b3b94faa DT	156	}
b3b94faa DT	157
b5d32bea SW	158	/**
	159	* inode_go_sync - Sync the dirty data and/or metadata for an inode glock
	160	* @gl: the glock protecting the inode
	161	*
	162	*/
	163
	164	static void inode_go_sync(struct gfs2_glock *gl)
	165	{
	166	struct gfs2_inode *ip = gl->gl_object;
009d8518	167	struct address_space *metamapping = gfs2_glock2aspace(gl);
3042a2cc SW	168	int error;
3042a2cc SW	169
b5d32bea SW	170	if (ip && !S_ISREG(ip->i_inode.i_mode))
b5d32bea SW	171	ip = NULL;
6bac243f SW	172	if (ip && test_and_clear_bit(GIF_SW_PAGED, &ip->i_flags))
	173	unmap_shared_mapping_range(ip->i_inode.i_mapping, 0, 0);
	174	if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags))
	175	return;
b5d32bea	176
6bac243f SW	177	BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE);
	178
	179	gfs2_log_flush(gl->gl_sbd, gl);
	180	filemap_fdatawrite(metamapping);
	181	if (ip) {
	182	struct address_space *mapping = ip->i_inode.i_mapping;
	183	filemap_fdatawrite(mapping);
	184	error = filemap_fdatawait(mapping);
	185	mapping_set_error(mapping, error);
b5d32bea	186	}
6bac243f SW	187	error = filemap_fdatawait(metamapping);
	188	mapping_set_error(metamapping, error);
	189	gfs2_ail_empty_gl(gl);
52fcd11c SW	190	/*
	191	* Writeback of the data mapping may cause the dirty flag to be set
	192	* so we have to clear it again here.
	193	*/
	194	smp_mb__before_clear_bit();
	195	clear_bit(GLF_DIRTY, &gl->gl_flags);
b5d32bea SW	196	}
b5d32bea SW	197
b3b94faa DT	198	/**
	199	* inode_go_inval - prepare a inode glock to be released
	200	* @gl: the glock
	201	* @flags:
6bac243f SW	202	*
	203	* Normally we invlidate everything, but if we are moving into
	204	* LM_ST_DEFERRED from LM_ST_SHARED or LM_ST_EXCLUSIVE then we
	205	* can keep hold of the metadata, since it won't have changed.
b3b94faa DT	206	*
	207	*/
	208
	209	static void inode_go_inval(struct gfs2_glock *gl, int flags)
	210	{
b004157a	211	struct gfs2_inode *ip = gl->gl_object;
b3b94faa	212
6bac243f SW	213	gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count));
	214
	215	if (flags & DIO_METADATA) {
009d8518	216	struct address_space *mapping = gfs2_glock2aspace(gl);
6bac243f	217	truncate_inode_pages(mapping, 0);
c65f7fb5	218	if (ip) {
b004157a	219	set_bit(GIF_INVALID, &ip->i_flags);
c65f7fb5 SW	220	forget_all_cached_acls(&ip->i_inode);
c65f7fb5 SW	221	}
b004157a SW	222	}
b004157a SW	223
1ce53368 BM	224	if (ip == GFS2_I(gl->gl_sbd->sd_rindex)) {
1ce53368 BM	225	gfs2_log_flush(gl->gl_sbd, NULL);
6bac243f	226	gl->gl_sbd->sd_rindex_uptodate = 0;
1ce53368	227	}
3cc3f710	228	if (ip && S_ISREG(ip->i_inode.i_mode))
b004157a	229	truncate_inode_pages(ip->i_inode.i_mapping, 0);
b3b94faa DT	230	}
	231
	232	/**
	233	* inode_go_demote_ok - Check to see if it's ok to unlock an inode glock
	234	* @gl: the glock
	235	*
	236	* Returns: 1 if it's ok
	237	*/
	238
97cc1025	239	static int inode_go_demote_ok(const struct gfs2_glock *gl)
b3b94faa DT	240	{
b3b94faa DT	241	struct gfs2_sbd *sdp = gl->gl_sbd;
bc015cb8 SW	242	struct gfs2_holder *gh;
bc015cb8 SW	243
97cc1025 SW	244	if (sdp->sd_jindex == gl->gl_object \|\| sdp->sd_rindex == gl->gl_object)
97cc1025 SW	245	return 0;
bc015cb8 SW	246
	247	if (!list_empty(&gl->gl_holders)) {
	248	gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list);
	249	if (gh->gh_list.next != &gl->gl_holders)
	250	return 0;
	251	}
	252
97cc1025	253	return 1;
b3b94faa DT	254	}
b3b94faa DT	255
d4b2cf1b SW	256	/**
	257	* gfs2_set_nlink - Set the inode's link count based on on-disk info
	258	* @inode: The inode in question
	259	* @nlink: The link count
	260	*
	261	* If the link count has hit zero, it must never be raised, whatever the
	262	* on-disk inode might say. When new struct inodes are created the link
	263	* count is set to 1, so that we can safely use this test even when reading
	264	* in on disk information for the first time.
	265	*/
	266
	267	static void gfs2_set_nlink(struct inode *inode, u32 nlink)
	268	{
	269	/*
	270	* We will need to review setting the nlink count here in the
	271	* light of the forthcoming ro bind mount work. This is a reminder
	272	* to do that.
	273	*/
	274	if ((inode->i_nlink != nlink) && (inode->i_nlink != 0)) {
	275	if (nlink == 0)
	276	clear_nlink(inode);
	277	else
	278	inode->i_nlink = nlink;
	279	}
	280	}
	281
	282	static int gfs2_dinode_in(struct gfs2_inode ip, const void buf)
	283	{
	284	const struct gfs2_dinode *str = buf;
	285	struct timespec atime;
	286	u16 height, depth;
	287
	288	if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr)))
	289	goto corrupt;
	290	ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino);
	291	ip->i_inode.i_mode = be32_to_cpu(str->di_mode);
	292	ip->i_inode.i_rdev = 0;
	293	switch (ip->i_inode.i_mode & S_IFMT) {
	294	case S_IFBLK:
	295	case S_IFCHR:
	296	ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major),
	297	be32_to_cpu(str->di_minor));
	298	break;
	299	};
	300
	301	ip->i_inode.i_uid = be32_to_cpu(str->di_uid);
	302	ip->i_inode.i_gid = be32_to_cpu(str->di_gid);
	303	gfs2_set_nlink(&ip->i_inode, be32_to_cpu(str->di_nlink));
	304	i_size_write(&ip->i_inode, be64_to_cpu(str->di_size));
	305	gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks));
	306	atime.tv_sec = be64_to_cpu(str->di_atime);
	307	atime.tv_nsec = be32_to_cpu(str->di_atime_nsec);
	308	if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0)
	309	ip->i_inode.i_atime = atime;
	310	ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime);
	311	ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec);
	312	ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime);
	313	ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec);
	314
	315	ip->i_goal = be64_to_cpu(str->di_goal_meta);
	316	ip->i_generation = be64_to_cpu(str->di_generation);
	317
	318	ip->i_diskflags = be32_to_cpu(str->di_flags);
	319	gfs2_set_inode_flags(&ip->i_inode);
320	height = be16_to_cpu(str->di_height);
321	if (unlikely(height > GFS2_MAX_META_HEIGHT))
322	goto corrupt;
323	ip->i_height = (u8)height;
324
325	depth = be16_to_cpu(str->di_depth);
326	if (unlikely(depth > GFS2_DIR_MAX_DEPTH))
327	goto corrupt;
328	ip->i_depth = (u8)depth;
329	ip->i_entries = be32_to_cpu(str->di_entries);
330
331	ip->i_eattr = be64_to_cpu(str->di_eattr);
332	if (S_ISREG(ip->i_inode.i_mode))
333	gfs2_set_aops(&ip->i_inode);
334
335	return 0;
336	corrupt:
337	gfs2_consist_inode(ip);
338	return -EIO;
339	}
340
341	/**
342	* gfs2_inode_refresh - Refresh the incore copy of the dinode
343	* @ip: The GFS2 inode
344	*
345	* Returns: errno
346	*/
347
348	int gfs2_inode_refresh(struct gfs2_inode *ip)
349	{
350	struct buffer_head *dibh;
351	int error;
352
353	error = gfs2_meta_inode_buffer(ip, &dibh);
354	if (error)
355	return error;
356
357	if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), dibh, GFS2_METATYPE_DI)) {
358	brelse(dibh);
359	return -EIO;
360	}
361
362	error = gfs2_dinode_in(ip, dibh->b_data);
363	brelse(dibh);
364	clear_bit(GIF_INVALID, &ip->i_flags);
365
366	return error;
367	}
368
b3b94faa DT	369	/**
	370	* inode_go_lock - operation done after an inode lock is locked by a process
	371	* @gl: the glock
	372	* @flags:
	373	*
	374	* Returns: errno
	375	*/
	376
	377	static int inode_go_lock(struct gfs2_holder *gh)
	378	{
	379	struct gfs2_glock *gl = gh->gh_gl;
813e0c46	380	struct gfs2_sbd *sdp = gl->gl_sbd;
5c676f6d	381	struct gfs2_inode *ip = gl->gl_object;
b3b94faa DT	382	int error = 0;
b3b94faa DT	383
091806ed	384	if (!ip \|\| (gh->gh_flags & GL_SKIP))
b3b94faa DT	385	return 0;
b3b94faa DT	386
bfded27b	387	if (test_bit(GIF_INVALID, &ip->i_flags)) {
b3b94faa DT	388	error = gfs2_inode_refresh(ip);
	389	if (error)
	390	return error;
b3b94faa DT	391	}
b3b94faa DT	392
383f01fb	393	if ((ip->i_diskflags & GFS2_DIF_TRUNC_IN_PROG) &&
b3b94faa	394	(gl->gl_state == LM_ST_EXCLUSIVE) &&
813e0c46 SW	395	(gh->gh_state == LM_ST_EXCLUSIVE)) {
	396	spin_lock(&sdp->sd_trunc_lock);
	397	if (list_empty(&ip->i_trunc_list))
	398	list_add(&sdp->sd_trunc_list, &ip->i_trunc_list);
	399	spin_unlock(&sdp->sd_trunc_lock);
	400	wake_up(&sdp->sd_quota_wait);
	401	return 1;
	402	}
b3b94faa DT	403
	404	return error;
	405	}
	406
6802e340 SW	407	/**
	408	* inode_go_dump - print information about an inode
	409	* @seq: The iterator
	410	* @ip: the inode
	411	*
	412	* Returns: 0 on success, -ENOBUFS when we run out of space
	413	*/
	414
	415	static int inode_go_dump(struct seq_file seq, const struct gfs2_glock gl)
	416	{
	417	const struct gfs2_inode *ip = gl->gl_object;
	418	if (ip == NULL)
	419	return 0;
a2e0f799	420	gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu\n",
6802e340 SW	421	(unsigned long long)ip->i_no_formal_ino,
6802e340 SW	422	(unsigned long long)ip->i_no_addr,
fa75cedc SW	423	IF2DT(ip->i_inode.i_mode), ip->i_flags,
fa75cedc SW	424	(unsigned int)ip->i_diskflags,
a2e0f799	425	(unsigned long long)i_size_read(&ip->i_inode));
6802e340 SW	426	return 0;
	427	}
	428
b3b94faa DT	429	/**
	430	* rgrp_go_lock - operation done after an rgrp lock is locked by
	431	* a first holder on this node.
	432	* @gl: the glock
	433	* @flags:
	434	*
	435	* Returns: errno
	436	*/
	437
	438	static int rgrp_go_lock(struct gfs2_holder *gh)
	439	{
5c676f6d	440	return gfs2_rgrp_bh_get(gh->gh_gl->gl_object);
b3b94faa DT	441	}
	442
	443	/**
	444	* rgrp_go_unlock - operation done before an rgrp lock is unlocked by
	445	* a last holder on this node.
	446	* @gl: the glock
	447	* @flags:
	448	*
	449	*/
	450
	451	static void rgrp_go_unlock(struct gfs2_holder *gh)
	452	{
5c676f6d	453	gfs2_rgrp_bh_put(gh->gh_gl->gl_object);
b3b94faa DT	454	}
	455
	456	/**
3042a2cc	457	* trans_go_sync - promote/demote the transaction glock
b3b94faa DT	458	* @gl: the glock
	459	* @state: the requested state
	460	* @flags:
	461	*
	462	*/
	463
3042a2cc	464	static void trans_go_sync(struct gfs2_glock *gl)
b3b94faa DT	465	{
	466	struct gfs2_sbd *sdp = gl->gl_sbd;
	467
	468	if (gl->gl_state != LM_ST_UNLOCKED &&
	469	test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
	470	gfs2_meta_syncfs(sdp);
	471	gfs2_log_shutdown(sdp);
	472	}
b3b94faa DT	473	}
	474
	475	/**
	476	* trans_go_xmote_bh - After promoting/demoting the transaction glock
	477	* @gl: the glock
	478	*
	479	*/
	480
6802e340	481	static int trans_go_xmote_bh(struct gfs2_glock gl, struct gfs2_holder gh)
b3b94faa DT	482	{
b3b94faa DT	483	struct gfs2_sbd *sdp = gl->gl_sbd;
feaa7bba	484	struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode);
5c676f6d	485	struct gfs2_glock *j_gl = ip->i_gl;
55167622	486	struct gfs2_log_header_host head;
b3b94faa DT	487	int error;
b3b94faa DT	488
6802e340	489	if (test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) {
1a14d3a6	490	j_gl->gl_ops->go_inval(j_gl, DIO_METADATA);
b3b94faa DT	491
	492	error = gfs2_find_jhead(sdp->sd_jdesc, &head);
	493	if (error)
	494	gfs2_consist(sdp);
	495	if (!(head.lh_flags & GFS2_LOG_HEAD_UNMOUNT))
	496	gfs2_consist(sdp);
	497
	498	/* Initialize some head of the log stuff */
	499	if (!test_bit(SDF_SHUTDOWN, &sdp->sd_flags)) {
	500	sdp->sd_log_sequence = head.lh_sequence + 1;
	501	gfs2_log_pointers_init(sdp, head.lh_blkno);
	502	}
	503	}
6802e340	504	return 0;
b3b94faa DT	505	}
b3b94faa DT	506
97cc1025 SW	507	/**
	508	* trans_go_demote_ok
	509	* @gl: the glock
	510	*
	511	* Always returns 0
	512	*/
	513
	514	static int trans_go_demote_ok(const struct gfs2_glock *gl)
	515	{
	516	return 0;
	517	}
	518
b94a170e BM	519	/**
	520	* iopen_go_callback - schedule the dcache entry for the inode to be deleted
	521	* @gl: the glock
	522	*
	523	* gl_spin lock is held while calling this
	524	*/
	525	static void iopen_go_callback(struct gfs2_glock *gl)
	526	{
	527	struct gfs2_inode ip = (struct gfs2_inode )gl->gl_object;
001e8e8d SW	528	struct gfs2_sbd *sdp = gl->gl_sbd;
	529
	530	if (sdp->sd_vfs->s_flags & MS_RDONLY)
	531	return;
b94a170e BM	532
b94a170e BM	533	if (gl->gl_demote_state == LM_ST_UNLOCKED &&
009d8518	534	gl->gl_state == LM_ST_SHARED && ip) {
b94a170e BM	535	gfs2_glock_hold(gl);
	536	if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
	537	gfs2_glock_put_nolock(gl);
	538	}
	539	}
	540
8fb4b536	541	const struct gfs2_glock_operations gfs2_meta_glops = {
ea67eedb	542	.go_type = LM_TYPE_META,
b3b94faa DT	543	};
b3b94faa DT	544
8fb4b536	545	const struct gfs2_glock_operations gfs2_inode_glops = {
3042a2cc	546	.go_xmote_th = inode_go_sync,
b3b94faa DT	547	.go_inval = inode_go_inval,
	548	.go_demote_ok = inode_go_demote_ok,
	549	.go_lock = inode_go_lock,
6802e340	550	.go_dump = inode_go_dump,
ea67eedb	551	.go_type = LM_TYPE_INODE,
6802e340	552	.go_min_hold_time = HZ / 5,
009d8518	553	.go_flags = GLOF_ASPACE,
b3b94faa DT	554	};
b3b94faa DT	555
8fb4b536	556	const struct gfs2_glock_operations gfs2_rgrp_glops = {
6bac243f SW	557	.go_xmote_th = rgrp_go_sync,
6bac243f SW	558	.go_inval = rgrp_go_inval,
b3b94faa DT	559	.go_lock = rgrp_go_lock,
b3b94faa DT	560	.go_unlock = rgrp_go_unlock,
09010978	561	.go_dump = gfs2_rgrp_dump,
ea67eedb	562	.go_type = LM_TYPE_RGRP,
6802e340	563	.go_min_hold_time = HZ / 5,
009d8518	564	.go_flags = GLOF_ASPACE,
b3b94faa DT	565	};
b3b94faa DT	566
8fb4b536	567	const struct gfs2_glock_operations gfs2_trans_glops = {
3042a2cc	568	.go_xmote_th = trans_go_sync,
b3b94faa	569	.go_xmote_bh = trans_go_xmote_bh,
97cc1025	570	.go_demote_ok = trans_go_demote_ok,
ea67eedb	571	.go_type = LM_TYPE_NONDISK,
b3b94faa DT	572	};
b3b94faa DT	573
8fb4b536	574	const struct gfs2_glock_operations gfs2_iopen_glops = {
ea67eedb	575	.go_type = LM_TYPE_IOPEN,
b94a170e	576	.go_callback = iopen_go_callback,
b3b94faa DT	577	};
b3b94faa DT	578
8fb4b536	579	const struct gfs2_glock_operations gfs2_flock_glops = {
ea67eedb	580	.go_type = LM_TYPE_FLOCK,
b3b94faa DT	581	};
b3b94faa DT	582
8fb4b536	583	const struct gfs2_glock_operations gfs2_nondisk_glops = {
ea67eedb	584	.go_type = LM_TYPE_NONDISK,
b3b94faa DT	585	};
b3b94faa DT	586
8fb4b536	587	const struct gfs2_glock_operations gfs2_quota_glops = {
ea67eedb	588	.go_type = LM_TYPE_QUOTA,
b3b94faa DT	589	};
b3b94faa DT	590
8fb4b536	591	const struct gfs2_glock_operations gfs2_journal_glops = {
ea67eedb	592	.go_type = LM_TYPE_JOURNAL,
b3b94faa DT	593	};
b3b94faa DT	594
64d576ba SW	595	const struct gfs2_glock_operations *gfs2_glops_list[] = {
	596	[LM_TYPE_META] = &gfs2_meta_glops,
	597	[LM_TYPE_INODE] = &gfs2_inode_glops,
	598	[LM_TYPE_RGRP] = &gfs2_rgrp_glops,
64d576ba SW	599	[LM_TYPE_IOPEN] = &gfs2_iopen_glops,
	600	[LM_TYPE_FLOCK] = &gfs2_flock_glops,
	601	[LM_TYPE_NONDISK] = &gfs2_nondisk_glops,
	602	[LM_TYPE_QUOTA] = &gfs2_quota_glops,
	603	[LM_TYPE_JOURNAL] = &gfs2_journal_glops,
	604	};
	605