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

// SPDX-License-Identifier: GPL-2.0
#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 __le32 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);
}

static int ext4_mmp_csum_verify(struct super_block *sb, struct mmp_struct *mmp)
{
	if (!ext4_has_metadata_csum(sb))
		return 1;

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

static void ext4_mmp_csum_set(struct super_block *sb, struct mmp_struct *mmp)
{
	if (!ext4_has_metadata_csum(sb))
		return;

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

/*
 * Write the MMP block using REQ_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);

	/*
	 * We protect against freezing so that we don't create dirty buffers
	 * on frozen filesystem.
	 */
	sb_start_write(sb);
	ext4_mmp_csum_set(sb, mmp);
	lock_buffer(bh);
	bh->b_end_io = end_buffer_write_sync;
	get_bh(bh);
	submit_bh(REQ_OP_WRITE, REQ_SYNC | REQ_META | REQ_PRIO, bh);
	wait_on_buffer(bh);
	sb_end_write(sb);
	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;
	int ret;

	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) {
			ret = -ENOMEM;
			goto warn_exit;
		}
	}

	get_bh(*bh);
	lock_buffer(*bh);
	(*bh)->b_end_io = end_buffer_read_sync;
	submit_bh(REQ_OP_READ, REQ_META | REQ_PRIO, *bh);
	wait_on_buffer(*bh);
	if (!buffer_uptodate(*bh)) {
		ret = -EIO;
		goto warn_exit;
	}
	mmp = (struct mmp_struct *)((*bh)->b_data);
	if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC) {
		ret = -EFSCORRUPTED;
		goto warn_exit;
	}
	if (!ext4_mmp_csum_verify(sb, mmp)) {
		ret = -EFSBADCRC;
		goto warn_exit;
	}
	return 0;
warn_exit:
	brelse(*bh);
	*bh = NULL;
	ext4_warning(sb, "Error %d while reading MMP block %llu",
		     ret, mmp_block);
	return ret;
}

/*
 * 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, "%s", msg);
	__ext4_warning(sb, function, line,
		       "MMP failure info: last update time: %llu, last update node: %.*s, last update device: %.*s",
		       (unsigned long long)le64_to_cpu(mmp->mmp_time),
		       (int)sizeof(mmp->mmp_nodename), mmp->mmp_nodename,
		       (int)sizeof(mmp->mmp_bdevname), 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);
	BUILD_BUG_ON(sizeof(mmp->mmp_bdevname) < BDEVNAME_SIZE);
	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.");
			goto exit_thread;
		}

		if (sb_rdonly(sb))
			break;

		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);
				goto exit_thread;
			}

			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");
				put_bh(bh_check);
				retval = -EBUSY;
				goto exit_thread;
			}
			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);

exit_thread:
	EXT4_SB(sb)->s_mmp_tsk = NULL;
	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 {
		new_seq = prandom_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");
		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(*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",
					     (int)sizeof(mmp->mmp_bdevname),
					     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
b2441318	1	// SPDX-License-Identifier: GPL-2.0
c5e06d10 JL	2	#include <linux/fs.h>
	3	#include <linux/random.h>
	4	#include <linux/buffer_head.h>
	5	#include <linux/utsname.h>
	6	#include <linux/kthread.h>
	7
	8	#include "ext4.h"
	9
5c359a47	10	/* Checksumming functions */
171a7f21	11	static __le32 ext4_mmp_csum(struct super_block sb, struct mmp_struct mmp)
5c359a47 DW	12	{
	13	struct ext4_sb_info *sbi = EXT4_SB(sb);
	14	int offset = offsetof(struct mmp_struct, mmp_checksum);
	15	__u32 csum;
	16
	17	csum = ext4_chksum(sbi, sbi->s_csum_seed, (char *)mmp, offset);
	18
	19	return cpu_to_le32(csum);
	20	}
	21
c197855e	22	static int ext4_mmp_csum_verify(struct super_block sb, struct mmp_struct mmp)
5c359a47	23	{
9aa5d32b	24	if (!ext4_has_metadata_csum(sb))
5c359a47 DW	25	return 1;
	26
	27	return mmp->mmp_checksum == ext4_mmp_csum(sb, mmp);
	28	}
	29
c197855e	30	static void ext4_mmp_csum_set(struct super_block sb, struct mmp_struct mmp)
5c359a47	31	{
9aa5d32b	32	if (!ext4_has_metadata_csum(sb))
5c359a47 DW	33	return;
	34
	35	mmp->mmp_checksum = ext4_mmp_csum(sb, mmp);
	36	}
	37
c5e06d10	38	/*
70fd7614	39	* Write the MMP block using REQ_SYNC to try to get the block on-disk
c5e06d10 JL	40	* faster.
c5e06d10 JL	41	*/
5c359a47	42	static int write_mmp_block(struct super_block sb, struct buffer_head bh)
c5e06d10	43	{
5c359a47 DW	44	struct mmp_struct mmp = (struct mmp_struct )(bh->b_data);
5c359a47 DW	45
8e8ad8a5 JK	46	/*
	47	* We protect against freezing so that we don't create dirty buffers
	48	* on frozen filesystem.
	49	*/
	50	sb_start_write(sb);
5c359a47	51	ext4_mmp_csum_set(sb, mmp);
c5e06d10 JL	52	lock_buffer(bh);
	53	bh->b_end_io = end_buffer_write_sync;
	54	get_bh(bh);
70fd7614	55	submit_bh(REQ_OP_WRITE, REQ_SYNC \| REQ_META \| REQ_PRIO, bh);
c5e06d10	56	wait_on_buffer(bh);
8e8ad8a5	57	sb_end_write(sb);
c5e06d10 JL	58	if (unlikely(!buffer_uptodate(bh)))
	59	return 1;
	60
	61	return 0;
	62	}
	63
	64	/*
	65	* Read the MMP block. It _must_ be read from disk and hence we clear the
	66	* uptodate flag on the buffer.
	67	*/
	68	static int read_mmp_block(struct super_block sb, struct buffer_head *bh,
	69	ext4_fsblk_t mmp_block)
	70	{
	71	struct mmp_struct *mmp;
98104468	72	int ret;
c5e06d10 JL	73
	74	if (*bh)
	75	clear_buffer_uptodate(*bh);
	76
	77	/* This would be sb_bread(sb, mmp_block), except we need to be sure
	78	* that the MD RAID device cache has been bypassed, and that the read
	79	* is not blocked in the elevator. */
98104468	80	if (!*bh) {
c5e06d10	81	*bh = sb_getblk(sb, mmp_block);
98104468 DC	82	if (!*bh) {
	83	ret = -ENOMEM;
	84	goto warn_exit;
c5e06d10 JL	85	}
c5e06d10 JL	86	}
98104468 DC	87
	88	get_bh(*bh);
	89	lock_buffer(*bh);
	90	(*bh)->b_end_io = end_buffer_read_sync;
70fd7614	91	submit_bh(REQ_OP_READ, REQ_META \| REQ_PRIO, *bh);
98104468 DC	92	wait_on_buffer(*bh);
98104468 DC	93	if (!buffer_uptodate(*bh)) {
98104468 DC	94	ret = -EIO;
98104468 DC	95	goto warn_exit;
c5e06d10	96	}
c5e06d10	97	mmp = (struct mmp_struct )((bh)->b_data);
03046886	98	if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC) {
6a797d27	99	ret = -EFSCORRUPTED;
03046886	100	goto warn_exit;
	101	}
	102	if (!ext4_mmp_csum_verify(sb, mmp)) {
6a797d27	103	ret = -EFSBADCRC;
03046886	104	goto warn_exit;
	105	}
	106	return 0;
98104468	107	warn_exit:
03046886	108	brelse(*bh);
03046886	109	*bh = NULL;
98104468 DC	110	ext4_warning(sb, "Error %d while reading MMP block %llu",
	111	ret, mmp_block);
	112	return ret;
c5e06d10 JL	113	}
	114
	115	/*
	116	* Dump as much information as possible to help the admin.
	117	*/
	118	void __dump_mmp_msg(struct super_block sb, struct mmp_struct mmp,
	119	const char function, unsigned int line, const char msg)
	120	{
da0b5e40	121	__ext4_warning(sb, function, line, "%s", msg);
c5e06d10	122	__ext4_warning(sb, function, line,
9da119de AD	123	"MMP failure info: last update time: %llu, last update node: %.s, last update device: %.s",
	124	(unsigned long long)le64_to_cpu(mmp->mmp_time),
	125	(int)sizeof(mmp->mmp_nodename), mmp->mmp_nodename,
	126	(int)sizeof(mmp->mmp_bdevname), mmp->mmp_bdevname);
c5e06d10 JL	127	}
	128
	129	/*
	130	* kmmpd will update the MMP sequence every s_mmp_update_interval seconds
	131	*/
	132	static int kmmpd(void *data)
	133	{
	134	struct super_block sb = ((struct mmpd_data ) data)->sb;
	135	struct buffer_head bh = ((struct mmpd_data ) data)->bh;
	136	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	137	struct mmp_struct *mmp;
	138	ext4_fsblk_t mmp_block;
	139	u32 seq = 0;
	140	unsigned long failed_writes = 0;
	141	int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval);
	142	unsigned mmp_check_interval;
	143	unsigned long last_update_time;
	144	unsigned long diff;
	145	int retval;
	146
	147	mmp_block = le64_to_cpu(es->s_mmp_block);
	148	mmp = (struct mmp_struct *)(bh->b_data);
	149	mmp->mmp_time = cpu_to_le64(get_seconds());
	150	/*
	151	* Start with the higher mmp_check_interval and reduce it if
	152	* the MMP block is being updated on time.
	153	*/
	154	mmp_check_interval = max(EXT4_MMP_CHECK_MULT * mmp_update_interval,
	155	EXT4_MMP_MIN_CHECK_INTERVAL);
	156	mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
9da119de	157	BUILD_BUG_ON(sizeof(mmp->mmp_bdevname) < BDEVNAME_SIZE);
c5e06d10 JL	158	bdevname(bh->b_bdev, mmp->mmp_bdevname);
c5e06d10 JL	159
215fc6af	160	memcpy(mmp->mmp_nodename, init_utsname()->nodename,
c5e06d10 JL	161	sizeof(mmp->mmp_nodename));
	162
	163	while (!kthread_should_stop()) {
	164	if (++seq > EXT4_MMP_SEQ_MAX)
	165	seq = 1;
	166
	167	mmp->mmp_seq = cpu_to_le32(seq);
	168	mmp->mmp_time = cpu_to_le64(get_seconds());
	169	last_update_time = jiffies;
	170
5c359a47	171	retval = write_mmp_block(sb, bh);
c5e06d10 JL	172	/*
	173	* Don't spew too many error messages. Print one every
	174	* (s_mmp_update_interval * 60) seconds.
	175	*/
bdfc230f NA	176	if (retval) {
	177	if ((failed_writes % 60) == 0)
	178	ext4_error(sb, "Error writing to MMP block");
c5e06d10 JL	179	failed_writes++;
	180	}
	181
	182	if (!(le32_to_cpu(es->s_feature_incompat) &
	183	EXT4_FEATURE_INCOMPAT_MMP)) {
	184	ext4_warning(sb, "kmmpd being stopped since MMP feature"
	185	" has been disabled.");
03046886	186	goto exit_thread;
c5e06d10 JL	187	}
c5e06d10 JL	188
9bed8748 TT	189	if (sb_rdonly(sb))
9bed8748 TT	190	break;
c5e06d10 JL	191
	192	diff = jiffies - last_update_time;
	193	if (diff < mmp_update_interval * HZ)
	194	schedule_timeout_interruptible(mmp_update_interval *
	195	HZ - diff);
	196
	197	/*
	198	* We need to make sure that more than mmp_check_interval
	199	* seconds have not passed since writing. If that has happened
	200	* we need to check if the MMP block is as we left it.
	201	*/
	202	diff = jiffies - last_update_time;
	203	if (diff > mmp_check_interval * HZ) {
	204	struct buffer_head *bh_check = NULL;
	205	struct mmp_struct *mmp_check;
	206
	207	retval = read_mmp_block(sb, &bh_check, mmp_block);
	208	if (retval) {
	209	ext4_error(sb, "error reading MMP data: %d",
	210	retval);
03046886	211	goto exit_thread;
c5e06d10 JL	212	}
	213
	214	mmp_check = (struct mmp_struct *)(bh_check->b_data);
	215	if (mmp->mmp_seq != mmp_check->mmp_seq \|\|
	216	memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename,
	217	sizeof(mmp->mmp_nodename))) {
	218	dump_mmp_msg(sb, mmp_check,
	219	"Error while updating MMP info. "
	220	"The filesystem seems to have been"
	221	" multiply mounted.");
	222	ext4_error(sb, "abort");
03046886	223	put_bh(bh_check);
	224	retval = -EBUSY;
	225	goto exit_thread;
c5e06d10 JL	226	}
	227	put_bh(bh_check);
	228	}
	229
	230	/*
	231	* Adjust the mmp_check_interval depending on how much time
	232	* it took for the MMP block to be written.
	233	*/
	234	mmp_check_interval = max(min(EXT4_MMP_CHECK_MULT * diff / HZ,
	235	EXT4_MMP_MAX_CHECK_INTERVAL),
	236	EXT4_MMP_MIN_CHECK_INTERVAL);
	237	mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval);
	238	}
	239
	240	/*
	241	* Unmount seems to be clean.
	242	*/
	243	mmp->mmp_seq = cpu_to_le32(EXT4_MMP_SEQ_CLEAN);
	244	mmp->mmp_time = cpu_to_le64(get_seconds());
	245
5c359a47	246	retval = write_mmp_block(sb, bh);
c5e06d10	247
03046886	248	exit_thread:
03046886	249	EXT4_SB(sb)->s_mmp_tsk = NULL;
c5e06d10 JL	250	kfree(data);
	251	brelse(bh);
	252	return retval;
	253	}
	254
	255	/*
	256	* Get a random new sequence number but make sure it is not greater than
	257	* EXT4_MMP_SEQ_MAX.
	258	*/
	259	static unsigned int mmp_new_seq(void)
	260	{
	261	u32 new_seq;
	262
	263	do {
dd1f723b	264	new_seq = prandom_u32();
c5e06d10 JL	265	} while (new_seq > EXT4_MMP_SEQ_MAX);
	266
	267	return new_seq;
	268	}
	269
	270	/*
	271	* Protect the filesystem from being mounted more than once.
	272	*/
	273	int ext4_multi_mount_protect(struct super_block *sb,
	274	ext4_fsblk_t mmp_block)
	275	{
	276	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
	277	struct buffer_head *bh = NULL;
	278	struct mmp_struct *mmp = NULL;
	279	struct mmpd_data *mmpd_data;
	280	u32 seq;
	281	unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval);
	282	unsigned int wait_time = 0;
	283	int retval;
	284
	285	if (mmp_block < le32_to_cpu(es->s_first_data_block) \|\|
	286	mmp_block >= ext4_blocks_count(es)) {
	287	ext4_warning(sb, "Invalid MMP block in superblock");
	288	goto failed;
	289	}
	290
	291	retval = read_mmp_block(sb, &bh, mmp_block);
	292	if (retval)
	293	goto failed;
	294
	295	mmp = (struct mmp_struct *)(bh->b_data);
	296
	297	if (mmp_check_interval < EXT4_MMP_MIN_CHECK_INTERVAL)
	298	mmp_check_interval = EXT4_MMP_MIN_CHECK_INTERVAL;
	299
	300	/*
	301	* If check_interval in MMP block is larger, use that instead of
	302	* update_interval from the superblock.
	303	*/
85d21650 SN	304	if (le16_to_cpu(mmp->mmp_check_interval) > mmp_check_interval)
85d21650 SN	305	mmp_check_interval = le16_to_cpu(mmp->mmp_check_interval);
c5e06d10 JL	306
	307	seq = le32_to_cpu(mmp->mmp_seq);
	308	if (seq == EXT4_MMP_SEQ_CLEAN)
	309	goto skip;
	310
	311	if (seq == EXT4_MMP_SEQ_FSCK) {
	312	dump_mmp_msg(sb, mmp, "fsck is running on the filesystem");
	313	goto failed;
	314	}
	315
	316	wait_time = min(mmp_check_interval * 2 + 1,
	317	mmp_check_interval + 60);
	318
	319	/* Print MMP interval if more than 20 secs. */
	320	if (wait_time > EXT4_MMP_MIN_CHECK_INTERVAL * 4)
	321	ext4_warning(sb, "MMP interval %u higher than expected, please"
	322	" wait.\n", wait_time * 2);
	323
	324	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
	325	ext4_warning(sb, "MMP startup interrupted, failing mount\n");
	326	goto failed;
	327	}
	328
	329	retval = read_mmp_block(sb, &bh, mmp_block);
	330	if (retval)
	331	goto failed;
	332	mmp = (struct mmp_struct *)(bh->b_data);
	333	if (seq != le32_to_cpu(mmp->mmp_seq)) {
	334	dump_mmp_msg(sb, mmp,
	335	"Device is already active on another node.");
	336	goto failed;
	337	}
	338
	339	skip:
	340	/*
	341	* write a new random sequence number.
	342	*/
f6f96fdb DW	343	seq = mmp_new_seq();
f6f96fdb DW	344	mmp->mmp_seq = cpu_to_le32(seq);
c5e06d10	345
5c359a47	346	retval = write_mmp_block(sb, bh);
c5e06d10 JL	347	if (retval)
	348	goto failed;
	349
	350	/*
	351	* wait for MMP interval and check mmp_seq.
	352	*/
	353	if (schedule_timeout_interruptible(HZ * wait_time) != 0) {
8d2ae1cb	354	ext4_warning(sb, "MMP startup interrupted, failing mount");
c5e06d10 JL	355	goto failed;
	356	}
	357
	358	retval = read_mmp_block(sb, &bh, mmp_block);
	359	if (retval)
	360	goto failed;
	361	mmp = (struct mmp_struct *)(bh->b_data);
	362	if (seq != le32_to_cpu(mmp->mmp_seq)) {
	363	dump_mmp_msg(sb, mmp,
	364	"Device is already active on another node.");
	365	goto failed;
	366	}
	367
d695a1be	368	mmpd_data = kmalloc(sizeof(*mmpd_data), GFP_KERNEL);
c5e06d10 JL	369	if (!mmpd_data) {
	370	ext4_warning(sb, "not enough memory for mmpd_data");
	371	goto failed;
	372	}
	373	mmpd_data->sb = sb;
	374	mmpd_data->bh = bh;
	375
	376	/*
	377	* Start a kernel thread to update the MMP block periodically.
	378	*/
9da119de AD	379	EXT4_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, mmpd_data, "kmmpd-%.*s",
9da119de AD	380	(int)sizeof(mmp->mmp_bdevname),
c5e06d10 JL	381	bdevname(bh->b_bdev,
	382	mmp->mmp_bdevname));
	383	if (IS_ERR(EXT4_SB(sb)->s_mmp_tsk)) {
	384	EXT4_SB(sb)->s_mmp_tsk = NULL;
	385	kfree(mmpd_data);
	386	ext4_warning(sb, "Unable to create kmmpd thread for %s.",
	387	sb->s_id);
	388	goto failed;
	389	}
	390
	391	return 0;
	392
	393	failed:
	394	brelse(bh);
	395	return 1;
	396	}
	397
	398