[linux-2.6-block.git] / fs / gfs2 / trans.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.
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
#include <linux/kallsyms.h>
#include <linux/gfs2_ondisk.h>

#include "gfs2.h"
#include "incore.h"
#include "glock.h"
#include "inode.h"
#include "log.h"
#include "lops.h"
#include "meta_io.h"
#include "trans.h"
#include "util.h"
#include "trace_gfs2.h"

int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
		     unsigned int revokes)
{
	struct gfs2_trans *tr;
	int error;

	BUG_ON(current->journal_info);
	BUG_ON(blocks == 0 && revokes == 0);

	if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
		return -EROFS;

	tr = kzalloc(sizeof(struct gfs2_trans), GFP_NOFS);
	if (!tr)
		return -ENOMEM;

	tr->tr_ip = (unsigned long)__builtin_return_address(0);
	tr->tr_blocks = blocks;
	tr->tr_revokes = revokes;
	tr->tr_reserved = 1;
	tr->tr_alloced = 1;
	if (blocks)
		tr->tr_reserved += 6 + blocks;
	if (revokes)
		tr->tr_reserved += gfs2_struct2blk(sdp, revokes,
						   sizeof(u64));
	INIT_LIST_HEAD(&tr->tr_databuf);
	INIT_LIST_HEAD(&tr->tr_buf);

	sb_start_intwrite(sdp->sd_vfs);

	error = gfs2_log_reserve(sdp, tr->tr_reserved);
	if (error)
		goto fail;

	current->journal_info = tr;

	return 0;

fail:
	sb_end_intwrite(sdp->sd_vfs);
	kfree(tr);

	return error;
}

static void gfs2_print_trans(const struct gfs2_trans *tr)
{
	pr_warn("Transaction created at: %pSR\n", (void *)tr->tr_ip);
	pr_warn("blocks=%u revokes=%u reserved=%u touched=%u\n",
		tr->tr_blocks, tr->tr_revokes, tr->tr_reserved, tr->tr_touched);
	pr_warn("Buf %u/%u Databuf %u/%u Revoke %u/%u\n",
		tr->tr_num_buf_new, tr->tr_num_buf_rm,
		tr->tr_num_databuf_new, tr->tr_num_databuf_rm,
		tr->tr_num_revoke, tr->tr_num_revoke_rm);
}

void gfs2_trans_end(struct gfs2_sbd *sdp)
{
	struct gfs2_trans *tr = current->journal_info;
	s64 nbuf;
	BUG_ON(!tr);
	current->journal_info = NULL;

	if (!tr->tr_touched) {
		gfs2_log_release(sdp, tr->tr_reserved);
		if (tr->tr_alloced)
			kfree(tr);
		sb_end_intwrite(sdp->sd_vfs);
		return;
	}

	nbuf = tr->tr_num_buf_new + tr->tr_num_databuf_new;
	nbuf -= tr->tr_num_buf_rm;
	nbuf -= tr->tr_num_databuf_rm;

	if (gfs2_assert_withdraw(sdp, (nbuf <= tr->tr_blocks) &&
				       (tr->tr_num_revoke <= tr->tr_revokes)))
		gfs2_print_trans(tr);

	gfs2_log_commit(sdp, tr);
	if (tr->tr_alloced && !tr->tr_attached)
			kfree(tr);
	up_read(&sdp->sd_log_flush_lock);

	if (sdp->sd_vfs->s_flags & MS_SYNCHRONOUS)
		gfs2_log_flush(sdp, NULL, NORMAL_FLUSH);
	sb_end_intwrite(sdp->sd_vfs);
}

static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl,
					       struct buffer_head *bh,
					       const struct gfs2_log_operations *lops)
{
	struct gfs2_bufdata *bd;

	bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL);
	bd->bd_bh = bh;
	bd->bd_gl = gl;
	bd->bd_ops = lops;
	INIT_LIST_HEAD(&bd->bd_list);
	bh->b_private = bd;
	return bd;
}

/**
 * gfs2_trans_add_data - Add a databuf to the transaction.
 * @gl: The inode glock associated with the buffer
 * @bh: The buffer to add
 *
 * This is used in two distinct cases:
 * i) In ordered write mode
 *    We put the data buffer on a list so that we can ensure that its
 *    synced to disk at the right time
 * ii) In journaled data mode
 *    We need to journal the data block in the same way as metadata in
 *    the functions above. The difference is that here we have a tag
 *    which is two __be64's being the block number (as per meta data)
 *    and a flag which says whether the data block needs escaping or
 *    not. This means we need a new log entry for each 251 or so data
 *    blocks, which isn't an enormous overhead but twice as much as
 *    for normal metadata blocks.
 */
void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh)
{
	struct gfs2_trans *tr = current->journal_info;
	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct address_space *mapping = bh->b_page->mapping;
	struct gfs2_inode *ip = GFS2_I(mapping->host);
	struct gfs2_bufdata *bd;

	if (!gfs2_is_jdata(ip)) {
		gfs2_ordered_add_inode(ip);
		return;
	}

	lock_buffer(bh);
	gfs2_log_lock(sdp);
	bd = bh->b_private;
	if (bd == NULL) {
		gfs2_log_unlock(sdp);
		unlock_buffer(bh);
		if (bh->b_private == NULL)
			bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops);
		lock_buffer(bh);
		gfs2_log_lock(sdp);
	}
	gfs2_assert(sdp, bd->bd_gl == gl);
	tr->tr_touched = 1;
	if (list_empty(&bd->bd_list)) {
		set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
		set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
		gfs2_pin(sdp, bd->bd_bh);
		tr->tr_num_databuf_new++;
		list_add_tail(&bd->bd_list, &tr->tr_databuf);
	}
	gfs2_log_unlock(sdp);
	unlock_buffer(bh);
}

static void meta_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd)
{
	struct gfs2_meta_header *mh;
	struct gfs2_trans *tr;

	tr = current->journal_info;
	tr->tr_touched = 1;
	if (!list_empty(&bd->bd_list))
		return;
	set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
	set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
	mh = (struct gfs2_meta_header *)bd->bd_bh->b_data;
	if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) {
		pr_err("Attempting to add uninitialised block to journal (inplace block=%lld)\n",
		       (unsigned long long)bd->bd_bh->b_blocknr);
		BUG();
	}
	gfs2_pin(sdp, bd->bd_bh);
	mh->__pad0 = cpu_to_be64(0);
	mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid);
	list_add(&bd->bd_list, &tr->tr_buf);
	tr->tr_num_buf_new++;
}

void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh)
{

	struct gfs2_sbd *sdp = gl->gl_sbd;
	struct gfs2_bufdata *bd;

	lock_buffer(bh);
	gfs2_log_lock(sdp);
	bd = bh->b_private;
	if (bd == NULL) {
		gfs2_log_unlock(sdp);
		unlock_buffer(bh);
		lock_page(bh->b_page);
		if (bh->b_private == NULL)
			bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops);
		unlock_page(bh->b_page);
		lock_buffer(bh);
		gfs2_log_lock(sdp);
	}
	gfs2_assert(sdp, bd->bd_gl == gl);
	meta_lo_add(sdp, bd);
	gfs2_log_unlock(sdp);
	unlock_buffer(bh);
}

void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd)
{
	struct gfs2_trans *tr = current->journal_info;

	BUG_ON(!list_empty(&bd->bd_list));
	gfs2_add_revoke(sdp, bd);
	tr->tr_touched = 1;
	tr->tr_num_revoke++;
}

void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
{
	struct gfs2_bufdata *bd, *tmp;
	struct gfs2_trans *tr = current->journal_info;
	unsigned int n = len;

	gfs2_log_lock(sdp);
	list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_list) {
		if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) {
			list_del_init(&bd->bd_list);
			gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
			sdp->sd_log_num_revoke--;
			kmem_cache_free(gfs2_bufdata_cachep, bd);
			tr->tr_num_revoke_rm++;
			if (--n == 0)
				break;
		}
	}
	gfs2_log_unlock(sdp);
}
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	*/
b3b94faa DT	9
d77d1b58 JP	10	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
d77d1b58 JP	11
b3b94faa DT	12	#include <linux/sched.h>
	13	#include <linux/slab.h>
	14	#include <linux/spinlock.h>
	15	#include <linux/completion.h>
	16	#include <linux/buffer_head.h>
d0dc80db	17	#include <linux/kallsyms.h>
f057f6cd	18	#include <linux/gfs2_ondisk.h>
b3b94faa DT	19
b3b94faa DT	20	#include "gfs2.h"
5c676f6d	21	#include "incore.h"
b3b94faa	22	#include "glock.h"
767f433f	23	#include "inode.h"
b3b94faa DT	24	#include "log.h"
	25	#include "lops.h"
	26	#include "meta_io.h"
	27	#include "trans.h"
5c676f6d	28	#include "util.h"
5e687eac	29	#include "trace_gfs2.h"
b3b94faa	30
d0dc80db SW	31	int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
d0dc80db SW	32	unsigned int revokes)
b3b94faa DT	33	{
	34	struct gfs2_trans *tr;
	35	int error;
	36
d0dc80db SW	37	BUG_ON(current->journal_info);
d0dc80db SW	38	BUG_ON(blocks == 0 && revokes == 0);
b3b94faa	39
a1c0643f SW	40	if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
	41	return -EROFS;
	42
f55ab26a	43	tr = kzalloc(sizeof(struct gfs2_trans), GFP_NOFS);
b3b94faa DT	44	if (!tr)
	45	return -ENOMEM;
	46
d0dc80db	47	tr->tr_ip = (unsigned long)__builtin_return_address(0);
b3b94faa DT	48	tr->tr_blocks = blocks;
	49	tr->tr_revokes = revokes;
	50	tr->tr_reserved = 1;
24972557	51	tr->tr_alloced = 1;
b3b94faa	52	if (blocks)
f4154ea0	53	tr->tr_reserved += 6 + blocks;
b3b94faa DT	54	if (revokes)
b3b94faa DT	55	tr->tr_reserved += gfs2_struct2blk(sdp, revokes,
cd915493	56	sizeof(u64));
d69a3c65 SW	57	INIT_LIST_HEAD(&tr->tr_databuf);
	58	INIT_LIST_HEAD(&tr->tr_buf);
	59
39263d5e	60	sb_start_intwrite(sdp->sd_vfs);
b3b94faa	61
b3b94faa DT	62	error = gfs2_log_reserve(sdp, tr->tr_reserved);
b3b94faa DT	63	if (error)
24972557	64	goto fail;
b3b94faa	65
5c676f6d	66	current->journal_info = tr;
b3b94faa DT	67
	68	return 0;
	69
24972557	70	fail:
39263d5e	71	sb_end_intwrite(sdp->sd_vfs);
b3b94faa DT	72	kfree(tr);
	73
	74	return error;
	75	}
	76
c50b91c4 SW	77	static void gfs2_print_trans(const struct gfs2_trans *tr)
c50b91c4 SW	78	{
d77d1b58 JP	79	pr_warn("Transaction created at: %pSR\n", (void *)tr->tr_ip);
	80	pr_warn("blocks=%u revokes=%u reserved=%u touched=%u\n",
	81	tr->tr_blocks, tr->tr_revokes, tr->tr_reserved, tr->tr_touched);
	82	pr_warn("Buf %u/%u Databuf %u/%u Revoke %u/%u\n",
	83	tr->tr_num_buf_new, tr->tr_num_buf_rm,
	84	tr->tr_num_databuf_new, tr->tr_num_databuf_rm,
	85	tr->tr_num_revoke, tr->tr_num_revoke_rm);
c50b91c4 SW	86	}
c50b91c4 SW	87
b3b94faa DT	88	void gfs2_trans_end(struct gfs2_sbd *sdp)
b3b94faa DT	89	{
f4154ea0	90	struct gfs2_trans *tr = current->journal_info;
c50b91c4	91	s64 nbuf;
f4154ea0	92	BUG_ON(!tr);
5c676f6d	93	current->journal_info = NULL;
b3b94faa	94
b3b94faa DT	95	if (!tr->tr_touched) {
b3b94faa DT	96	gfs2_log_release(sdp, tr->tr_reserved);
24972557	97	if (tr->tr_alloced)
d8348de0	98	kfree(tr);
39263d5e	99	sb_end_intwrite(sdp->sd_vfs);
b3b94faa DT	100	return;
	101	}
	102
c50b91c4 SW	103	nbuf = tr->tr_num_buf_new + tr->tr_num_databuf_new;
	104	nbuf -= tr->tr_num_buf_rm;
	105	nbuf -= tr->tr_num_databuf_rm;
	106
	107	if (gfs2_assert_withdraw(sdp, (nbuf <= tr->tr_blocks) &&
	108	(tr->tr_num_revoke <= tr->tr_revokes)))
	109	gfs2_print_trans(tr);
b3b94faa DT	110
b3b94faa DT	111	gfs2_log_commit(sdp, tr);
24972557	112	if (tr->tr_alloced && !tr->tr_attached)
16ca9412	113	kfree(tr);
16ca9412	114	up_read(&sdp->sd_log_flush_lock);
b3b94faa DT	115
b3b94faa DT	116	if (sdp->sd_vfs->s_flags & MS_SYNCHRONOUS)
24972557	117	gfs2_log_flush(sdp, NULL, NORMAL_FLUSH);
39263d5e	118	sb_end_intwrite(sdp->sd_vfs);
b3b94faa DT	119	}
b3b94faa DT	120
c76c4d96 SW	121	static struct gfs2_bufdata gfs2_alloc_bufdata(struct gfs2_glock gl,
	122	struct buffer_head *bh,
	123	const struct gfs2_log_operations *lops)
	124	{
	125	struct gfs2_bufdata *bd;
	126
	127	bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS \| __GFP_NOFAIL);
	128	bd->bd_bh = bh;
	129	bd->bd_gl = gl;
	130	bd->bd_ops = lops;
	131	INIT_LIST_HEAD(&bd->bd_list);
	132	bh->b_private = bd;
	133	return bd;
	134	}
	135
b3b94faa	136	/**
45138990 SW	137	* gfs2_trans_add_data - Add a databuf to the transaction.
	138	* @gl: The inode glock associated with the buffer
	139	* @bh: The buffer to add
b3b94faa	140	*
767f433f SW	141	* This is used in two distinct cases:
	142	* i) In ordered write mode
	143	* We put the data buffer on a list so that we can ensure that its
	144	* synced to disk at the right time
	145	* ii) In journaled data mode
	146	* We need to journal the data block in the same way as metadata in
	147	* the functions above. The difference is that here we have a tag
	148	* which is two __be64's being the block number (as per meta data)
	149	* and a flag which says whether the data block needs escaping or
	150	* not. This means we need a new log entry for each 251 or so data
	151	* blocks, which isn't an enormous overhead but twice as much as
	152	* for normal metadata blocks.
b3b94faa	153	*/
45138990	154	void gfs2_trans_add_data(struct gfs2_glock gl, struct buffer_head bh)
767f433f SW	155	{
767f433f SW	156	struct gfs2_trans *tr = current->journal_info;
45138990 SW	157	struct gfs2_sbd *sdp = gl->gl_sbd;
45138990 SW	158	struct address_space *mapping = bh->b_page->mapping;
767f433f	159	struct gfs2_inode *ip = GFS2_I(mapping->host);
45138990	160	struct gfs2_bufdata *bd;
767f433f	161
45138990 SW	162	if (!gfs2_is_jdata(ip)) {
45138990 SW	163	gfs2_ordered_add_inode(ip);
767f433f	164	return;
767f433f	165	}
b3b94faa	166
96e5d1d3 BM	167	lock_buffer(bh);
96e5d1d3 BM	168	gfs2_log_lock(sdp);
5c676f6d	169	bd = bh->b_private;
c76c4d96	170	if (bd == NULL) {
96e5d1d3 BM	171	gfs2_log_unlock(sdp);
96e5d1d3 BM	172	unlock_buffer(bh);
c76c4d96 SW	173	if (bh->b_private == NULL)
c76c4d96 SW	174	bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops);
96e5d1d3 BM	175	lock_buffer(bh);
96e5d1d3 BM	176	gfs2_log_lock(sdp);
b3b94faa	177	}
c76c4d96	178	gfs2_assert(sdp, bd->bd_gl == gl);
45138990 SW	179	tr->tr_touched = 1;
	180	if (list_empty(&bd->bd_list)) {
	181	set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
	182	set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
	183	gfs2_pin(sdp, bd->bd_bh);
	184	tr->tr_num_databuf_new++;
d69a3c65	185	list_add_tail(&bd->bd_list, &tr->tr_databuf);
45138990	186	}
96e5d1d3 BM	187	gfs2_log_unlock(sdp);
96e5d1d3 BM	188	unlock_buffer(bh);
b3b94faa DT	189	}
b3b94faa DT	190
767f433f	191	static void meta_lo_add(struct gfs2_sbd sdp, struct gfs2_bufdata bd)
350a9b0a	192	{
767f433f SW	193	struct gfs2_meta_header *mh;
	194	struct gfs2_trans *tr;
	195
	196	tr = current->journal_info;
	197	tr->tr_touched = 1;
	198	if (!list_empty(&bd->bd_list))
	199	return;
	200	set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
	201	set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
	202	mh = (struct gfs2_meta_header *)bd->bd_bh->b_data;
	203	if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) {
d77d1b58	204	pr_err("Attempting to add uninitialised block to journal (inplace block=%lld)\n",
767f433f SW	205	(unsigned long long)bd->bd_bh->b_blocknr);
	206	BUG();
	207	}
	208	gfs2_pin(sdp, bd->bd_bh);
	209	mh->__pad0 = cpu_to_be64(0);
	210	mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid);
d69a3c65	211	list_add(&bd->bd_list, &tr->tr_buf);
767f433f	212	tr->tr_num_buf_new++;
350a9b0a SW	213	}
	214
	215	void gfs2_trans_add_meta(struct gfs2_glock gl, struct buffer_head bh)
	216	{
767f433f SW	217
	218	struct gfs2_sbd *sdp = gl->gl_sbd;
	219	struct gfs2_bufdata *bd;
	220
	221	lock_buffer(bh);
	222	gfs2_log_lock(sdp);
	223	bd = bh->b_private;
c76c4d96	224	if (bd == NULL) {
767f433f SW	225	gfs2_log_unlock(sdp);
767f433f SW	226	unlock_buffer(bh);
c76c4d96 SW	227	lock_page(bh->b_page);
	228	if (bh->b_private == NULL)
	229	bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops);
	230	unlock_page(bh->b_page);
767f433f SW	231	lock_buffer(bh);
	232	gfs2_log_lock(sdp);
	233	}
c76c4d96	234	gfs2_assert(sdp, bd->bd_gl == gl);
767f433f SW	235	meta_lo_add(sdp, bd);
	236	gfs2_log_unlock(sdp);
	237	unlock_buffer(bh);
350a9b0a SW	238	}
350a9b0a SW	239
1ad38c43	240	void gfs2_trans_add_revoke(struct gfs2_sbd sdp, struct gfs2_bufdata bd)
b3b94faa	241	{
75f2b879 SW	242	struct gfs2_trans *tr = current->journal_info;
75f2b879 SW	243
c0752aa7	244	BUG_ON(!list_empty(&bd->bd_list));
5d054964	245	gfs2_add_revoke(sdp, bd);
75f2b879 SW	246	tr->tr_touched = 1;
75f2b879 SW	247	tr->tr_num_revoke++;
b3b94faa DT	248	}
b3b94faa DT	249
5731be53	250	void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
b3b94faa	251	{
5731be53 SW	252	struct gfs2_bufdata bd, tmp;
	253	struct gfs2_trans *tr = current->journal_info;
	254	unsigned int n = len;
b3b94faa DT	255
b3b94faa DT	256	gfs2_log_lock(sdp);
c0752aa7	257	list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_list) {
5731be53	258	if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) {
c0752aa7	259	list_del_init(&bd->bd_list);
b3b94faa DT	260	gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
b3b94faa DT	261	sdp->sd_log_num_revoke--;
5731be53 SW	262	kmem_cache_free(gfs2_bufdata_cachep, bd);
	263	tr->tr_num_revoke_rm++;
	264	if (--n == 0)
	265	break;
b3b94faa DT	266	}
b3b94faa DT	267	}
b3b94faa	268	gfs2_log_unlock(sdp);
b3b94faa DT	269	}
b3b94faa DT	270