[mirror_ubuntu-bionic-kernel.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;
	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);
	gfs2_holder_init(sdp->sd_trans_gl, LM_ST_SHARED, 0, &tr->tr_t_gh);

	error = gfs2_glock_nq(&tr->tr_t_gh);
	if (error)
		goto fail_holder_uninit;

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

	current->journal_info = tr;

	return 0;

fail_gunlock:
	gfs2_glock_dq(&tr->tr_t_gh);

fail_holder_uninit:
	sb_end_intwrite(sdp->sd_vfs);
	gfs2_holder_uninit(&tr->tr_t_gh);
	kfree(tr);

	return error;
}

/**
 * gfs2_log_release - Release a given number of log blocks
 * @sdp: The GFS2 superblock
 * @blks: The number of blocks
 *
 */

static void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
{

	atomic_add(blks, &sdp->sd_log_blks_free);
	trace_gfs2_log_blocks(sdp, blks);
	gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <=
				  sdp->sd_jdesc->jd_blocks);
	up_read(&sdp->sd_log_flush_lock);
}

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_t_gh.gh_gl) {
			gfs2_glock_dq(&tr->tr_t_gh);
			gfs2_holder_uninit(&tr->tr_t_gh);
			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_t_gh.gh_gl) {
		gfs2_glock_dq(&tr->tr_t_gh);
		gfs2_holder_uninit(&tr->tr_t_gh);
		if (!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);
	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
	1	/*
	2	* Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved.
	3	* Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
	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
	7	* of the GNU General Public License version 2.
	8	*/
	9
	10	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
	11
	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>
	17	#include <linux/kallsyms.h>
	18	#include <linux/gfs2_ondisk.h>
	19
	20	#include "gfs2.h"
	21	#include "incore.h"
	22	#include "glock.h"
	23	#include "inode.h"
	24	#include "log.h"
	25	#include "lops.h"
	26	#include "meta_io.h"
	27	#include "trans.h"
	28	#include "util.h"
	29	#include "trace_gfs2.h"
	30
	31	int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks,
	32	unsigned int revokes)
	33	{
	34	struct gfs2_trans *tr;
	35	int error;
	36
	37	BUG_ON(current->journal_info);
	38	BUG_ON(blocks == 0 && revokes == 0);
	39
	40	if (!test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags))
	41	return -EROFS;
	42
	43	tr = kzalloc(sizeof(struct gfs2_trans), GFP_NOFS);
	44	if (!tr)
	45	return -ENOMEM;
	46
	47	tr->tr_ip = (unsigned long)__builtin_return_address(0);
	48	tr->tr_blocks = blocks;
	49	tr->tr_revokes = revokes;
	50	tr->tr_reserved = 1;
	51	if (blocks)
	52	tr->tr_reserved += 6 + blocks;
	53	if (revokes)
	54	tr->tr_reserved += gfs2_struct2blk(sdp, revokes,
	55	sizeof(u64));
	56	INIT_LIST_HEAD(&tr->tr_databuf);
	57	INIT_LIST_HEAD(&tr->tr_buf);
	58
	59	sb_start_intwrite(sdp->sd_vfs);
	60	gfs2_holder_init(sdp->sd_trans_gl, LM_ST_SHARED, 0, &tr->tr_t_gh);
	61
	62	error = gfs2_glock_nq(&tr->tr_t_gh);
	63	if (error)
	64	goto fail_holder_uninit;
	65
	66	error = gfs2_log_reserve(sdp, tr->tr_reserved);
	67	if (error)
	68	goto fail_gunlock;
	69
	70	current->journal_info = tr;
	71
	72	return 0;
	73
	74	fail_gunlock:
	75	gfs2_glock_dq(&tr->tr_t_gh);
	76
	77	fail_holder_uninit:
	78	sb_end_intwrite(sdp->sd_vfs);
	79	gfs2_holder_uninit(&tr->tr_t_gh);
	80	kfree(tr);
	81
	82	return error;
	83	}
	84
	85	/**
	86	* gfs2_log_release - Release a given number of log blocks
	87	* @sdp: The GFS2 superblock
	88	* @blks: The number of blocks
	89	*
	90	*/
	91
	92	static void gfs2_log_release(struct gfs2_sbd *sdp, unsigned int blks)
	93	{
	94
	95	atomic_add(blks, &sdp->sd_log_blks_free);
	96	trace_gfs2_log_blocks(sdp, blks);
	97	gfs2_assert_withdraw(sdp, atomic_read(&sdp->sd_log_blks_free) <=
	98	sdp->sd_jdesc->jd_blocks);
	99	up_read(&sdp->sd_log_flush_lock);
	100	}
	101
	102	static void gfs2_print_trans(const struct gfs2_trans *tr)
	103	{
	104	pr_warn("Transaction created at: %pSR\n", (void *)tr->tr_ip);
	105	pr_warn("blocks=%u revokes=%u reserved=%u touched=%u\n",
	106	tr->tr_blocks, tr->tr_revokes, tr->tr_reserved, tr->tr_touched);
	107	pr_warn("Buf %u/%u Databuf %u/%u Revoke %u/%u\n",
	108	tr->tr_num_buf_new, tr->tr_num_buf_rm,
	109	tr->tr_num_databuf_new, tr->tr_num_databuf_rm,
	110	tr->tr_num_revoke, tr->tr_num_revoke_rm);
	111	}
	112
	113	void gfs2_trans_end(struct gfs2_sbd *sdp)
	114	{
	115	struct gfs2_trans *tr = current->journal_info;
	116	s64 nbuf;
	117	BUG_ON(!tr);
	118	current->journal_info = NULL;
	119
	120	if (!tr->tr_touched) {
	121	gfs2_log_release(sdp, tr->tr_reserved);
	122	if (tr->tr_t_gh.gh_gl) {
	123	gfs2_glock_dq(&tr->tr_t_gh);
	124	gfs2_holder_uninit(&tr->tr_t_gh);
	125	kfree(tr);
	126	}
	127	sb_end_intwrite(sdp->sd_vfs);
	128	return;
	129	}
	130
	131	nbuf = tr->tr_num_buf_new + tr->tr_num_databuf_new;
	132	nbuf -= tr->tr_num_buf_rm;
	133	nbuf -= tr->tr_num_databuf_rm;
	134
	135	if (gfs2_assert_withdraw(sdp, (nbuf <= tr->tr_blocks) &&
	136	(tr->tr_num_revoke <= tr->tr_revokes)))
	137	gfs2_print_trans(tr);
	138
	139	gfs2_log_commit(sdp, tr);
	140	if (tr->tr_t_gh.gh_gl) {
	141	gfs2_glock_dq(&tr->tr_t_gh);
	142	gfs2_holder_uninit(&tr->tr_t_gh);
	143	if (!tr->tr_attached)
	144	kfree(tr);
	145	}
	146	up_read(&sdp->sd_log_flush_lock);
	147
	148	if (sdp->sd_vfs->s_flags & MS_SYNCHRONOUS)
	149	gfs2_log_flush(sdp, NULL);
	150	sb_end_intwrite(sdp->sd_vfs);
	151	}
	152
	153	static struct gfs2_bufdata gfs2_alloc_bufdata(struct gfs2_glock gl,
	154	struct buffer_head *bh,
	155	const struct gfs2_log_operations *lops)
	156	{
	157	struct gfs2_bufdata *bd;
	158
	159	bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS \| __GFP_NOFAIL);
	160	bd->bd_bh = bh;
	161	bd->bd_gl = gl;
	162	bd->bd_ops = lops;
	163	INIT_LIST_HEAD(&bd->bd_list);
	164	bh->b_private = bd;
	165	return bd;
	166	}
	167
	168	/**
	169	* gfs2_trans_add_data - Add a databuf to the transaction.
	170	* @gl: The inode glock associated with the buffer
	171	* @bh: The buffer to add
	172	*
	173	* This is used in two distinct cases:
	174	* i) In ordered write mode
	175	* We put the data buffer on a list so that we can ensure that its
	176	* synced to disk at the right time
	177	* ii) In journaled data mode
	178	* We need to journal the data block in the same way as metadata in
	179	* the functions above. The difference is that here we have a tag
	180	* which is two __be64's being the block number (as per meta data)
	181	* and a flag which says whether the data block needs escaping or
	182	* not. This means we need a new log entry for each 251 or so data
	183	* blocks, which isn't an enormous overhead but twice as much as
	184	* for normal metadata blocks.
	185	*/
	186	void gfs2_trans_add_data(struct gfs2_glock gl, struct buffer_head bh)
	187	{
	188	struct gfs2_trans *tr = current->journal_info;
	189	struct gfs2_sbd *sdp = gl->gl_sbd;
	190	struct address_space *mapping = bh->b_page->mapping;
	191	struct gfs2_inode *ip = GFS2_I(mapping->host);
	192	struct gfs2_bufdata *bd;
	193
	194	if (!gfs2_is_jdata(ip)) {
	195	gfs2_ordered_add_inode(ip);
	196	return;
	197	}
	198
	199	lock_buffer(bh);
	200	gfs2_log_lock(sdp);
	201	bd = bh->b_private;
	202	if (bd == NULL) {
	203	gfs2_log_unlock(sdp);
	204	unlock_buffer(bh);
	205	if (bh->b_private == NULL)
	206	bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops);
	207	lock_buffer(bh);
	208	gfs2_log_lock(sdp);
	209	}
	210	gfs2_assert(sdp, bd->bd_gl == gl);
	211	tr->tr_touched = 1;
	212	if (list_empty(&bd->bd_list)) {
	213	set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
	214	set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
	215	gfs2_pin(sdp, bd->bd_bh);
	216	tr->tr_num_databuf_new++;
	217	list_add_tail(&bd->bd_list, &tr->tr_databuf);
	218	}
	219	gfs2_log_unlock(sdp);
	220	unlock_buffer(bh);
	221	}
	222
	223	static void meta_lo_add(struct gfs2_sbd sdp, struct gfs2_bufdata bd)
	224	{
	225	struct gfs2_meta_header *mh;
	226	struct gfs2_trans *tr;
	227
	228	tr = current->journal_info;
	229	tr->tr_touched = 1;
	230	if (!list_empty(&bd->bd_list))
	231	return;
	232	set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags);
	233	set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags);
	234	mh = (struct gfs2_meta_header *)bd->bd_bh->b_data;
	235	if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) {
	236	pr_err("Attempting to add uninitialised block to journal (inplace block=%lld)\n",
	237	(unsigned long long)bd->bd_bh->b_blocknr);
	238	BUG();
	239	}
	240	gfs2_pin(sdp, bd->bd_bh);
	241	mh->__pad0 = cpu_to_be64(0);
	242	mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid);
	243	list_add(&bd->bd_list, &tr->tr_buf);
	244	tr->tr_num_buf_new++;
	245	}
	246
	247	void gfs2_trans_add_meta(struct gfs2_glock gl, struct buffer_head bh)
	248	{
	249
	250	struct gfs2_sbd *sdp = gl->gl_sbd;
	251	struct gfs2_bufdata *bd;
	252
	253	lock_buffer(bh);
	254	gfs2_log_lock(sdp);
	255	bd = bh->b_private;
	256	if (bd == NULL) {
	257	gfs2_log_unlock(sdp);
	258	unlock_buffer(bh);
	259	lock_page(bh->b_page);
	260	if (bh->b_private == NULL)
	261	bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops);
	262	unlock_page(bh->b_page);
	263	lock_buffer(bh);
	264	gfs2_log_lock(sdp);
	265	}
	266	gfs2_assert(sdp, bd->bd_gl == gl);
	267	meta_lo_add(sdp, bd);
	268	gfs2_log_unlock(sdp);
	269	unlock_buffer(bh);
	270	}
	271
	272	void gfs2_trans_add_revoke(struct gfs2_sbd sdp, struct gfs2_bufdata bd)
	273	{
	274	struct gfs2_trans *tr = current->journal_info;
	275
	276	BUG_ON(!list_empty(&bd->bd_list));
	277	gfs2_add_revoke(sdp, bd);
	278	tr->tr_touched = 1;
	279	tr->tr_num_revoke++;
	280	}
	281
	282	void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len)
	283	{
	284	struct gfs2_bufdata bd, tmp;
	285	struct gfs2_trans *tr = current->journal_info;
	286	unsigned int n = len;
	287
	288	gfs2_log_lock(sdp);
	289	list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_list) {
	290	if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) {
	291	list_del_init(&bd->bd_list);
	292	gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke);
	293	sdp->sd_log_num_revoke--;
	294	kmem_cache_free(gfs2_bufdata_cachep, bd);
	295	tr->tr_num_revoke_rm++;
	296	if (--n == 0)
	297	break;
	298	}
	299	}
	300	gfs2_log_unlock(sdp);
	301	}
	302