[mirror_ubuntu-bionic-kernel.git] / fs / ext4 / mmp.c

#include <linux/fs.h>
#include <linux/random.h>
#include <linux/buffer_head.h>
#include <linux/utsname.h>
#include <linux/kthread.h>

#include "ext4.h"

/* Checksumming functions */
static __u32 ext4_mmp_csum(struct super_block *sb, struct mmp_struct *mmp)
{
	struct ext4_sb_info *sbi = EXT4_SB(sb);
	int offset = offsetof(struct mmp_struct, mmp_checksum);
	__u32 csum;

	csum = ext4_chksum(sbi, sbi->s_csum_seed, (char *)mmp, offset);

	return cpu_to_le32(csum);
}

int ext4_mmp_csum_verify(struct super_block *sb, struct mmp_struct *mmp)
{
	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
				       EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
		return 1;

	return mmp->mmp_checksum == ext4_mmp_csum(sb, mmp);
}

void ext4_mmp_csum_set(struct super_block *sb, struct mmp_struct *mmp)
{
	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
				       EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
		return;

	mmp->mmp_checksum = ext4_mmp_csum(sb, mmp);
}

/*
 * Write the MMP block using WRITE_SYNC to try to get the block on-disk
 * faster.
 */
static int write_mmp_block(struct super_block *sb, struct buffer_head *bh)
{
	struct mmp_struct *mmp = (struct mmp_struct *)(bh->b_data);

	ext4_mmp_csum_set(sb, mmp);
	mark_buffer_dirty(bh);
	lock_buffer(bh);
	bh->b_end_io = end_buffer_write_sync;
	get_bh(bh);
	submit_bh(WRITE_SYNC, bh);
	wait_on_buffer(bh);
	if (unlikely(!buffer_uptodate(bh)))
		return 1;

	return 0;
}

/*
 * Read the MMP block. It _must_ be read from disk and hence we clear the
 * uptodate flag on the buffer.
 */
static int read_mmp_block(struct super_block *sb, struct buffer_head **bh,
			  ext4_fsblk_t mmp_block)
{
	struct mmp_struct *mmp;

	if (*bh)
		clear_buffer_uptodate(*bh);

	/* This would be sb_bread(sb, mmp_block), except we need to be sure
	 * that the MD RAID device cache has been bypassed, and that the read
	 * is not blocked in the elevator. */
	if (!*bh)
		*bh = sb_getblk(sb, mmp_block);
	if (*bh) {
		get_bh(*bh);
		lock_buffer(*bh);
		(*bh)->b_end_io = end_buffer_read_sync;
		submit_bh(READ_SYNC, *bh);
		wait_on_buffer(*bh);
		if (!buffer_uptodate(*bh)) {
			brelse(*bh);
			*bh = NULL;
		}
	}
	if (!*bh) {
		ext4_warning(sb, "Error while reading MMP block %llu",
			     mmp_block);
		return -EIO;
	}

	mmp = (struct mmp_struct *)((*bh)->b_data);
	if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC ||
	    !ext4_mmp_csum_verify(sb, mmp))
		return -EINVAL;

	return 0;
}

/*
 * Dump as much information as possible to help the admin.
 */
void __dump_mmp_msg(struct super_block *sb, struct mmp_struct *mmp,
		    const char *function, unsigned int line, const char *msg)
{
	__ext4_warning(sb, function, line, msg);
	__ext4_warning(sb, function, line,
		       "MMP failure info: last update time: %llu, last update "
		       "node: %s, last update device: %s\n",
		       (long long unsigned int) le64_to_cpu(mmp->mmp_time),
		       mmp->mmp_nodename, mmp->mmp_bdevname);
}

/*
 * kmmpd will update the MMP sequence every s_mmp_update_interval seconds
 */
static int kmmpd(void *data)
{
	struct super_block *sb = ((struct mmpd_data *) data)->sb;
	struct buffer_head *bh = ((struct mmpd_data *) data)->bh;
	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	struct mmp_struct *mmp;
	ext4_fsblk_t mmp_block;
	u32 seq = 0;
	unsigned long failed_writes = 0;
	int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval);
	unsigned mmp_check_interval;
	unsigned long last_update_time;
	unsigned long diff;
	int retval;

	mmp_block = le64_to_cpu(es->s_mmp_block);
	mmp = (struct mmp_struct *)(bh->b_data);
	mmp->mmp_time = cpu_to_le64(get_seconds());
	/*
	 * Start with the higher mmp_check_interval and reduce it if
	 * the MMP block is being updated on time.
	 */
	mmp_check_interval = max(EXT4_MMP_CHECK_MULT * mmp_update_interval,
				 EXT4_MMP_MIN_CHECK_INTERVAL);
	mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
	bdevname(bh->b_bdev, mmp->mmp_bdevname);

	memcpy(mmp->mmp_nodename, init_utsname()->nodename,
	       sizeof(mmp->mmp_nodename));

	while (!kthread_should_stop()) {
		if (++seq > EXT4_MMP_SEQ_MAX)
			seq = 1;

		mmp->mmp_seq = cpu_to_le32(seq);
		mmp->mmp_time = cpu_to_le64(get_seconds());
		last_update_time = jiffies;

		retval = write_mmp_block(sb, bh);
		/*
		 * Don't spew too many error messages. Print one every
		 * (s_mmp_update_interval * 60) seconds.
		 */
		if (retval) {
			if ((failed_writes % 60) == 0)
				ext4_error(sb, "Error writing to MMP block");
			failed_writes++;
		}

		if (!(le32_to_cpu(es->s_feature_incompat) &
		    EXT4_FEATURE_INCOMPAT_MMP)) {
			ext4_warning(sb, "kmmpd being stopped since MMP feature"
				     " has been disabled.");
			EXT4_SB(sb)->s_mmp_tsk = NULL;
			goto failed;
		}

		if (sb->s_flags & MS_RDONLY) {
			ext4_warning(sb, "kmmpd being stopped since filesystem "
				     "has been remounted as readonly.");
			EXT4_SB(sb)->s_mmp_tsk = NULL;
			goto failed;
		}

		diff = jiffies - last_update_time;
		if (diff < mmp_update_interval * HZ)
			schedule_timeout_interruptible(mmp_update_interval *
						       HZ - diff);

		/*
		 * We need to make sure that more than mmp_check_interval
		 * seconds have not passed since writing. If that has happened
		 * we need to check if the MMP block is as we left it.
		 */
		diff = jiffies - last_update_time;
		if (diff > mmp_check_interval * HZ) {
			struct buffer_head *bh_check = NULL;
			struct mmp_struct *mmp_check;

			retval = read_mmp_block(sb, &bh_check, mmp_block);
			if (retval) {
				ext4_error(sb, "error reading MMP data: %d",
					   retval);

				EXT4_SB(sb)->s_mmp_tsk = NULL;
				goto failed;
			}

			mmp_check = (struct mmp_struct *)(bh_check->b_data);
			if (mmp->mmp_seq != mmp_check->mmp_seq ||
			    memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename,
				   sizeof(mmp->mmp_nodename))) {
				dump_mmp_msg(sb, mmp_check,
					     "Error while updating MMP info. "
					     "The filesystem seems to have been"
					     " multiply mounted.");
				ext4_error(sb, "abort");
				goto failed;
			}
			put_bh(bh_check);
		}

		 /*
		 * Adjust the mmp_check_interval depending on how much time
		 * it took for the MMP block to be written.
		 */
		mmp_check_interval = max(min(EXT4_MMP_CHECK_MULT * diff / HZ,
					     EXT4_MMP_MAX_CHECK_INTERVAL),
					 EXT4_MMP_MIN_CHECK_INTERVAL);
		mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
	}

	/*
	 * Unmount seems to be clean.
	 */
	mmp->mmp_seq = cpu_to_le32(EXT4_MMP_SEQ_CLEAN);
	mmp->mmp_time = cpu_to_le64(get_seconds());

	retval = write_mmp_block(sb, bh);

failed:
	kfree(data);
	brelse(bh);
	return retval;
}

/*
 * Get a random new sequence number but make sure it is not greater than
 * EXT4_MMP_SEQ_MAX.
 */
static unsigned int mmp_new_seq(void)
{
	u32 new_seq;

	do {
		get_random_bytes(&new_seq, sizeof(u32));
	} while (new_seq > EXT4_MMP_SEQ_MAX);

	return new_seq;
}

/*
 * Protect the filesystem from being mounted more than once.
 */
int ext4_multi_mount_protect(struct super_block *sb,
				    ext4_fsblk_t mmp_block)
{
	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	struct buffer_head *bh = NULL;
	struct mmp_struct *mmp = NULL;
	struct mmpd_data *mmpd_data;
	u32 seq;
	unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval);
	unsigned int wait_time = 0;
	int retval;

	if (mmp_block < le32_to_cpu(es->s_first_data_block) ||
	    mmp_block >= ext4_blocks_count(es)) {
		ext4_warning(sb, "Invalid MMP block in superblock");
		goto failed;
	}

	retval = read_mmp_block(sb, &bh, mmp_block);
	if (retval)
		goto failed;

	mmp = (struct mmp_struct *)(bh->b_data);

	if (mmp_check_interval < EXT4_MMP_MIN_CHECK_INTERVAL)
		mmp_check_interval = EXT4_MMP_MIN_CHECK_INTERVAL;

	/*
	 * If check_interval in MMP block is larger, use that instead of
	 * update_interval from the superblock.
	 */
	if (le16_to_cpu(mmp->mmp_check_interval) > mmp_check_interval)
		mmp_check_interval = le16_to_cpu(mmp->mmp_check_interval);

	seq = le32_to_cpu(mmp->mmp_seq);
	if (seq == EXT4_MMP_SEQ_CLEAN)
		goto skip;

	if (seq == EXT4_MMP_SEQ_FSCK) {
		dump_mmp_msg(sb, mmp, "fsck is running on the filesystem");
		goto failed;
	}

	wait_time = min(mmp_check_interval * 2 + 1,
			mmp_check_interval + 60);

	/* Print MMP interval if more than 20 secs. */
	if (wait_time > EXT4_MMP_MIN_CHECK_INTERVAL * 4)
		ext4_warning(sb, "MMP interval %u higher than expected, please"
			     " wait.\n", wait_time * 2);

	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
		ext4_warning(sb, "MMP startup interrupted, failing mount\n");
		goto failed;
	}

	retval = read_mmp_block(sb, &bh, mmp_block);
	if (retval)
		goto failed;
	mmp = (struct mmp_struct *)(bh->b_data);
	if (seq != le32_to_cpu(mmp->mmp_seq)) {
		dump_mmp_msg(sb, mmp,
			     "Device is already active on another node.");
		goto failed;
	}

skip:
	/*
	 * write a new random sequence number.
	 */
	seq = mmp_new_seq();
	mmp->mmp_seq = cpu_to_le32(seq);

	retval = write_mmp_block(sb, bh);
	if (retval)
		goto failed;

	/*
	 * wait for MMP interval and check mmp_seq.
	 */
	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
		ext4_warning(sb, "MMP startup interrupted, failing mount\n");
		goto failed;
	}

	retval = read_mmp_block(sb, &bh, mmp_block);
	if (retval)
		goto failed;
	mmp = (struct mmp_struct *)(bh->b_data);
	if (seq != le32_to_cpu(mmp->mmp_seq)) {
		dump_mmp_msg(sb, mmp,
			     "Device is already active on another node.");
		goto failed;
	}

	mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
	if (!mmpd_data) {
		ext4_warning(sb, "not enough memory for mmpd_data");
		goto failed;
	}
	mmpd_data->sb = sb;
	mmpd_data->bh = bh;

	/*
	 * Start a kernel thread to update the MMP block periodically.
	 */
	EXT4_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, mmpd_data, "kmmpd-%s",
					     bdevname(bh->b_bdev,
						      mmp->mmp_bdevname));
	if (IS_ERR(EXT4_SB(sb)->s_mmp_tsk)) {
		EXT4_SB(sb)->s_mmp_tsk = NULL;
		kfree(mmpd_data);
		ext4_warning(sb, "Unable to create kmmpd thread for %s.",
			     sb->s_id);
		goto failed;
	}

	return 0;

failed:
	brelse(bh);
	return 1;
}
Commit	Line	Data
c5e06d10 JL	1	#include <linux/fs.h>
	2	#include <linux/random.h>
	3	#include <linux/buffer_head.h>
	4	#include <linux/utsname.h>
	5	#include <linux/kthread.h>
	6
	7	#include "ext4.h"
	8
5c359a47 DW	9	/* Checksumming functions */
	10	static __u32 ext4_mmp_csum(struct super_block sb, struct mmp_struct mmp)
	11	{
	12	struct ext4_sb_info *sbi = EXT4_SB(sb);
	13	int offset = offsetof(struct mmp_struct, mmp_checksum);
	14	__u32 csum;
	15
	16	csum = ext4_chksum(sbi, sbi->s_csum_seed, (char *)mmp, offset);
	17
	18	return cpu_to_le32(csum);
	19	}
	20
	21	int ext4_mmp_csum_verify(struct super_block sb, struct mmp_struct mmp)
	22	{
	23	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
	24	EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
	25	return 1;
	26
	27	return mmp->mmp_checksum == ext4_mmp_csum(sb, mmp);
	28	}
	29
	30	void ext4_mmp_csum_set(struct super_block sb, struct mmp_struct mmp)
	31	{
	32	if (!EXT4_HAS_RO_COMPAT_FEATURE(sb,
	33	EXT4_FEATURE_RO_COMPAT_METADATA_CSUM))
	34	return;
	35
	36	mmp->mmp_checksum = ext4_mmp_csum(sb, mmp);
	37	}
	38
c5e06d10 JL	39	/*
	40	* Write the MMP block using WRITE_SYNC to try to get the block on-disk
	41	* faster.
	42	*/
5c359a47	43	static int write_mmp_block(struct super_block sb, struct buffer_head bh)
c5e06d10	44	{
5c359a47 DW	45	struct mmp_struct mmp = (struct mmp_struct )(bh->b_data);
	46
	47	ext4_mmp_csum_set(sb, mmp);
c5e06d10 JL	48	mark_buffer_dirty(bh);
	49	lock_buffer(bh);
	50	bh->b_end_io = end_buffer_write_sync;
	51	get_bh(bh);
	52	submit_bh(WRITE_SYNC, bh);
	53	wait_on_buffer(bh);
	54	if (unlikely(!buffer_uptodate(bh)))
	55	return 1;
	56
	57	return 0;
	58	}
	59
	60	/*
	61	* Read the MMP block. It _must_ be read from disk and hence we clear the
	62	* uptodate flag on the buffer.
	63	*/
	64	static int read_mmp_block(struct super_block sb, struct buffer_head *bh,
	65	ext4_fsblk_t mmp_block)
	66	{
	67	struct mmp_struct *mmp;
	68
	69	if (*bh)
	70	clear_buffer_uptodate(*bh);
	71
	72	/* This would be sb_bread(sb, mmp_block), except we need to be sure
	73	* that the MD RAID device cache has been bypassed, and that the read
	74	* is not blocked in the elevator. */
	75	if (!*bh)
	76	*bh = sb_getblk(sb, mmp_block);
	77	if (*bh) {
	78	get_bh(*bh);
	79	lock_buffer(*bh);
	80	(*bh)->b_end_io = end_buffer_read_sync;
	81	submit_bh(READ_SYNC, *bh);
	82	wait_on_buffer(*bh);
	83	if (!buffer_uptodate(*bh)) {
	84	brelse(*bh);
	85	*bh = NULL;
	86	}
	87	}
	88	if (!*bh) {
	89	ext4_warning(sb, "Error while reading MMP block %llu",
	90	mmp_block);
	91	return -EIO;
	92	}
	93
	94	mmp = (struct mmp_struct )((bh)->b_data);
5c359a47 DW	95	if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC \|\|
5c359a47 DW	96	!ext4_mmp_csum_verify(sb, mmp))
c5e06d10 JL	97	return -EINVAL;
	98
	99	return 0;
	100	}
	101
	102	/*
	103	* Dump as much information as possible to help the admin.
	104	*/
	105	void __dump_mmp_msg(struct super_block sb, struct mmp_struct mmp,
	106	const char function, unsigned int line, const char msg)
	107	{
	108	__ext4_warning(sb, function, line, msg);
	109	__ext4_warning(sb, function, line,
	110	"MMP failure info: last update time: %llu, last update "
	111	"node: %s, last update device: %s\n",
	112	(long long unsigned int) le64_to_cpu(mmp->mmp_time),
	113	mmp->mmp_nodename, mmp->mmp_bdevname);
	114	}
	115
	116	/*
	117	* kmmpd will update the MMP sequence every s_mmp_update_interval seconds
	118	*/
	119	static int kmmpd(void *data)
	120	{
	121	struct super_block sb = ((struct mmpd_data ) data)->sb;
	122	struct buffer_head bh = ((struct mmpd_data ) data)->bh;
	123	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	124	struct mmp_struct *mmp;
	125	ext4_fsblk_t mmp_block;
	126	u32 seq = 0;
	127	unsigned long failed_writes = 0;
	128	int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval);
	129	unsigned mmp_check_interval;
	130	unsigned long last_update_time;
	131	unsigned long diff;
	132	int retval;
	133
	134	mmp_block = le64_to_cpu(es->s_mmp_block);
	135	mmp = (struct mmp_struct *)(bh->b_data);
	136	mmp->mmp_time = cpu_to_le64(get_seconds());
	137	/*
	138	* Start with the higher mmp_check_interval and reduce it if
	139	* the MMP block is being updated on time.
	140	*/
	141	mmp_check_interval = max(EXT4_MMP_CHECK_MULT * mmp_update_interval,
	142	EXT4_MMP_MIN_CHECK_INTERVAL);
	143	mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
	144	bdevname(bh->b_bdev, mmp->mmp_bdevname);
	145
215fc6af	146	memcpy(mmp->mmp_nodename, init_utsname()->nodename,
c5e06d10 JL	147	sizeof(mmp->mmp_nodename));
	148
	149	while (!kthread_should_stop()) {
	150	if (++seq > EXT4_MMP_SEQ_MAX)
	151	seq = 1;
	152
	153	mmp->mmp_seq = cpu_to_le32(seq);
	154	mmp->mmp_time = cpu_to_le64(get_seconds());
	155	last_update_time = jiffies;
	156
5c359a47	157	retval = write_mmp_block(sb, bh);
c5e06d10 JL	158	/*
	159	* Don't spew too many error messages. Print one every
	160	* (s_mmp_update_interval * 60) seconds.
	161	*/
bdfc230f NA	162	if (retval) {
	163	if ((failed_writes % 60) == 0)
	164	ext4_error(sb, "Error writing to MMP block");
c5e06d10 JL	165	failed_writes++;
	166	}
	167
	168	if (!(le32_to_cpu(es->s_feature_incompat) &
	169	EXT4_FEATURE_INCOMPAT_MMP)) {
	170	ext4_warning(sb, "kmmpd being stopped since MMP feature"
	171	" has been disabled.");
	172	EXT4_SB(sb)->s_mmp_tsk = NULL;
	173	goto failed;
	174	}
	175
	176	if (sb->s_flags & MS_RDONLY) {
	177	ext4_warning(sb, "kmmpd being stopped since filesystem "
	178	"has been remounted as readonly.");
	179	EXT4_SB(sb)->s_mmp_tsk = NULL;
	180	goto failed;
	181	}
	182
	183	diff = jiffies - last_update_time;
	184	if (diff < mmp_update_interval * HZ)
	185	schedule_timeout_interruptible(mmp_update_interval *
	186	HZ - diff);
	187
	188	/*
	189	* We need to make sure that more than mmp_check_interval
	190	* seconds have not passed since writing. If that has happened
	191	* we need to check if the MMP block is as we left it.
	192	*/
	193	diff = jiffies - last_update_time;
	194	if (diff > mmp_check_interval * HZ) {
	195	struct buffer_head *bh_check = NULL;
	196	struct mmp_struct *mmp_check;
	197
	198	retval = read_mmp_block(sb, &bh_check, mmp_block);
	199	if (retval) {
	200	ext4_error(sb, "error reading MMP data: %d",
	201	retval);
	202
	203	EXT4_SB(sb)->s_mmp_tsk = NULL;
	204	goto failed;
	205	}
	206
	207	mmp_check = (struct mmp_struct *)(bh_check->b_data);
	208	if (mmp->mmp_seq != mmp_check->mmp_seq \|\|
	209	memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename,
	210	sizeof(mmp->mmp_nodename))) {
	211	dump_mmp_msg(sb, mmp_check,
	212	"Error while updating MMP info. "
	213	"The filesystem seems to have been"
	214	" multiply mounted.");
	215	ext4_error(sb, "abort");
	216	goto failed;
	217	}
	218	put_bh(bh_check);
	219	}
	220
	221	/*
	222	* Adjust the mmp_check_interval depending on how much time
	223	* it took for the MMP block to be written.
	224	*/
	225	mmp_check_interval = max(min(EXT4_MMP_CHECK_MULT * diff / HZ,
	226	EXT4_MMP_MAX_CHECK_INTERVAL),
	227	EXT4_MMP_MIN_CHECK_INTERVAL);
	228	mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
229	}
230
231	/*
232	* Unmount seems to be clean.
233	*/
234	mmp->mmp_seq = cpu_to_le32(EXT4_MMP_SEQ_CLEAN);
235	mmp->mmp_time = cpu_to_le64(get_seconds());
236
5c359a47	237	retval = write_mmp_block(sb, bh);
c5e06d10 JL	238
	239	failed:
	240	kfree(data);
	241	brelse(bh);
	242	return retval;
	243	}
	244
	245	/*
	246	* Get a random new sequence number but make sure it is not greater than
	247	* EXT4_MMP_SEQ_MAX.
	248	*/
	249	static unsigned int mmp_new_seq(void)
	250	{
	251	u32 new_seq;
	252
	253	do {
	254	get_random_bytes(&new_seq, sizeof(u32));
	255	} while (new_seq > EXT4_MMP_SEQ_MAX);
	256
	257	return new_seq;
	258	}
	259
	260	/*
	261	* Protect the filesystem from being mounted more than once.
	262	*/
	263	int ext4_multi_mount_protect(struct super_block *sb,
	264	ext4_fsblk_t mmp_block)
	265	{
	266	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	267	struct buffer_head *bh = NULL;
	268	struct mmp_struct *mmp = NULL;
	269	struct mmpd_data *mmpd_data;
	270	u32 seq;
	271	unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval);
	272	unsigned int wait_time = 0;
	273	int retval;
	274
	275	if (mmp_block < le32_to_cpu(es->s_first_data_block) \|\|
	276	mmp_block >= ext4_blocks_count(es)) {
	277	ext4_warning(sb, "Invalid MMP block in superblock");
	278	goto failed;
	279	}
	280
	281	retval = read_mmp_block(sb, &bh, mmp_block);
	282	if (retval)
	283	goto failed;
	284
	285	mmp = (struct mmp_struct *)(bh->b_data);
	286
	287	if (mmp_check_interval < EXT4_MMP_MIN_CHECK_INTERVAL)
	288	mmp_check_interval = EXT4_MMP_MIN_CHECK_INTERVAL;
	289
	290	/*
	291	* If check_interval in MMP block is larger, use that instead of
	292	* update_interval from the superblock.
	293	*/
85d21650 SN	294	if (le16_to_cpu(mmp->mmp_check_interval) > mmp_check_interval)
85d21650 SN	295	mmp_check_interval = le16_to_cpu(mmp->mmp_check_interval);
c5e06d10 JL	296
	297	seq = le32_to_cpu(mmp->mmp_seq);
	298	if (seq == EXT4_MMP_SEQ_CLEAN)
	299	goto skip;
	300
	301	if (seq == EXT4_MMP_SEQ_FSCK) {
	302	dump_mmp_msg(sb, mmp, "fsck is running on the filesystem");
	303	goto failed;
	304	}
	305
	306	wait_time = min(mmp_check_interval * 2 + 1,
	307	mmp_check_interval + 60);
	308
	309	/* Print MMP interval if more than 20 secs. */
	310	if (wait_time > EXT4_MMP_MIN_CHECK_INTERVAL * 4)
	311	ext4_warning(sb, "MMP interval %u higher than expected, please"
	312	" wait.\n", wait_time * 2);
	313
	314	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
	315	ext4_warning(sb, "MMP startup interrupted, failing mount\n");
	316	goto failed;
	317	}
	318
	319	retval = read_mmp_block(sb, &bh, mmp_block);
	320	if (retval)
	321	goto failed;
	322	mmp = (struct mmp_struct *)(bh->b_data);
	323	if (seq != le32_to_cpu(mmp->mmp_seq)) {
	324	dump_mmp_msg(sb, mmp,
	325	"Device is already active on another node.");
	326	goto failed;
	327	}
	328
	329	skip:
	330	/*
	331	* write a new random sequence number.
	332	*/
f6f96fdb DW	333	seq = mmp_new_seq();
f6f96fdb DW	334	mmp->mmp_seq = cpu_to_le32(seq);
c5e06d10	335
5c359a47	336	retval = write_mmp_block(sb, bh);
c5e06d10 JL	337	if (retval)
	338	goto failed;
	339
	340	/*
	341	* wait for MMP interval and check mmp_seq.
	342	*/
	343	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
	344	ext4_warning(sb, "MMP startup interrupted, failing mount\n");
	345	goto failed;
	346	}
	347
	348	retval = read_mmp_block(sb, &bh, mmp_block);
	349	if (retval)
	350	goto failed;
	351	mmp = (struct mmp_struct *)(bh->b_data);
	352	if (seq != le32_to_cpu(mmp->mmp_seq)) {
	353	dump_mmp_msg(sb, mmp,
	354	"Device is already active on another node.");
	355	goto failed;
	356	}
	357
	358	mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL);
	359	if (!mmpd_data) {
	360	ext4_warning(sb, "not enough memory for mmpd_data");
	361	goto failed;
	362	}
	363	mmpd_data->sb = sb;
	364	mmpd_data->bh = bh;
	365
	366	/*
	367	* Start a kernel thread to update the MMP block periodically.
	368	*/
	369	EXT4_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, mmpd_data, "kmmpd-%s",
	370	bdevname(bh->b_bdev,
	371	mmp->mmp_bdevname));
	372	if (IS_ERR(EXT4_SB(sb)->s_mmp_tsk)) {
	373	EXT4_SB(sb)->s_mmp_tsk = NULL;
	374	kfree(mmpd_data);
	375	ext4_warning(sb, "Unable to create kmmpd thread for %s.",
	376	sb->s_id);
	377	goto failed;
	378	}
	379
	380	return 0;
	381
	382	failed:
	383	brelse(bh);
	384	return 1;
	385	}
	386
	387