path: root/drivers/block/zloop.c

// SPDX-License-Identifier: GPL-2.0-only
/*
 * Copyright (c) 2025, Christoph Hellwig.
 * Copyright (c) 2025, Western Digital Corporation or its affiliates.
 *
 * Zoned Loop Device driver - exports a zoned block device using one file per
 * zone as backing storage.
 */
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/module.h>
#include <linux/blk-mq.h>
#include <linux/blkzoned.h>
#include <linux/pagemap.h>
#include <linux/miscdevice.h>
#include <linux/falloc.h>
#include <linux/mutex.h>
#include <linux/parser.h>
#include <linux/seq_file.h>

/*
 * Options for adding (and removing) a device.
 */
enum {
	ZLOOP_OPT_ERR			= 0,
	ZLOOP_OPT_ID			= (1 << 0),
	ZLOOP_OPT_CAPACITY		= (1 << 1),
	ZLOOP_OPT_ZONE_SIZE		= (1 << 2),
	ZLOOP_OPT_ZONE_CAPACITY		= (1 << 3),
	ZLOOP_OPT_NR_CONV_ZONES		= (1 << 4),
	ZLOOP_OPT_BASE_DIR		= (1 << 5),
	ZLOOP_OPT_NR_QUEUES		= (1 << 6),
	ZLOOP_OPT_QUEUE_DEPTH		= (1 << 7),
	ZLOOP_OPT_BUFFERED_IO		= (1 << 8),
	ZLOOP_OPT_ZONE_APPEND		= (1 << 9),
	ZLOOP_OPT_ORDERED_ZONE_APPEND	= (1 << 10),
};

static const match_table_t zloop_opt_tokens = {
	{ ZLOOP_OPT_ID,			"id=%d"	},
	{ ZLOOP_OPT_CAPACITY,		"capacity_mb=%u"	},
	{ ZLOOP_OPT_ZONE_SIZE,		"zone_size_mb=%u"	},
	{ ZLOOP_OPT_ZONE_CAPACITY,	"zone_capacity_mb=%u"	},
	{ ZLOOP_OPT_NR_CONV_ZONES,	"conv_zones=%u"		},
	{ ZLOOP_OPT_BASE_DIR,		"base_dir=%s"		},
	{ ZLOOP_OPT_NR_QUEUES,		"nr_queues=%u"		},
	{ ZLOOP_OPT_QUEUE_DEPTH,	"queue_depth=%u"	},
	{ ZLOOP_OPT_BUFFERED_IO,	"buffered_io"		},
	{ ZLOOP_OPT_ZONE_APPEND,	"zone_append=%u"	},
	{ ZLOOP_OPT_ORDERED_ZONE_APPEND, "ordered_zone_append"	},
	{ ZLOOP_OPT_ERR,		NULL			}
};

/* Default values for the "add" operation. */
#define ZLOOP_DEF_ID			-1
#define ZLOOP_DEF_ZONE_SIZE		((256ULL * SZ_1M) >> SECTOR_SHIFT)
#define ZLOOP_DEF_NR_ZONES		64
#define ZLOOP_DEF_NR_CONV_ZONES		8
#define ZLOOP_DEF_BASE_DIR		"/var/local/zloop"
#define ZLOOP_DEF_NR_QUEUES		1
#define ZLOOP_DEF_QUEUE_DEPTH		128
#define ZLOOP_DEF_BUFFERED_IO		false
#define ZLOOP_DEF_ZONE_APPEND		true
#define ZLOOP_DEF_ORDERED_ZONE_APPEND	false

/* Arbitrary limit on the zone size (16GB). */
#define ZLOOP_MAX_ZONE_SIZE_MB		16384

struct zloop_options {
	unsigned int		mask;
	int			id;
	sector_t		capacity;
	sector_t		zone_size;
	sector_t		zone_capacity;
	unsigned int		nr_conv_zones;
	char			*base_dir;
	unsigned int		nr_queues;
	unsigned int		queue_depth;
	bool			buffered_io;
	bool			zone_append;
	bool			ordered_zone_append;
};

/*
 * Device states.
 */
enum {
	Zlo_creating = 0,
	Zlo_live,
	Zlo_deleting,
};

enum zloop_zone_flags {
	ZLOOP_ZONE_CONV = 0,
	ZLOOP_ZONE_SEQ_ERROR,
};

struct zloop_zone {
	struct file		*file;

	unsigned long		flags;
	struct mutex		lock;
	spinlock_t		wp_lock;
	enum blk_zone_cond	cond;
	sector_t		start;
	sector_t		wp;

	gfp_t			old_gfp_mask;
};

struct zloop_device {
	unsigned int		id;
	unsigned int		state;

	struct blk_mq_tag_set	tag_set;
	struct gendisk		*disk;

	struct workqueue_struct *workqueue;
	bool			buffered_io;
	bool			zone_append;
	bool			ordered_zone_append;

	const char		*base_dir;
	struct file		*data_dir;

	unsigned int		zone_shift;
	sector_t		zone_size;
	sector_t		zone_capacity;
	unsigned int		nr_zones;
	unsigned int		nr_conv_zones;
	unsigned int		block_size;

	struct zloop_zone	zones[] __counted_by(nr_zones);
};

struct zloop_cmd {
	struct work_struct	work;
	atomic_t		ref;
	sector_t		sector;
	sector_t		nr_sectors;
	long			ret;
	struct kiocb		iocb;
	struct bio_vec		*bvec;
};

static DEFINE_IDR(zloop_index_idr);
static DEFINE_MUTEX(zloop_ctl_mutex);

static unsigned int rq_zone_no(struct request *rq)
{
	struct zloop_device *zlo = rq->q->queuedata;

	return blk_rq_pos(rq) >> zlo->zone_shift;
}

static int zloop_update_seq_zone(struct zloop_device *zlo, unsigned int zone_no)
{
	struct zloop_zone *zone = &zlo->zones[zone_no];
	struct kstat stat;
	sector_t file_sectors;
	unsigned long flags;
	int ret;

	lockdep_assert_held(&zone->lock);

	ret = vfs_getattr(&zone->file->f_path, &stat, STATX_SIZE, 0);
	if (ret < 0) {
		pr_err("Failed to get zone %u file stat (err=%d)\n",
		       zone_no, ret);
		set_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags);
		return ret;
	}

	file_sectors = stat.size >> SECTOR_SHIFT;
	if (file_sectors > zlo->zone_capacity) {
		pr_err("Zone %u file too large (%llu sectors > %llu)\n",
		       zone_no, file_sectors, zlo->zone_capacity);
		return -EINVAL;
	}

	if (file_sectors & ((zlo->block_size >> SECTOR_SHIFT) - 1)) {
		pr_err("Zone %u file size not aligned to block size %u\n",
		       zone_no, zlo->block_size);
		return -EINVAL;
	}

	spin_lock_irqsave(&zone->wp_lock, flags);
	if (!file_sectors) {
		zone->cond = BLK_ZONE_COND_EMPTY;
		zone->wp = zone->start;
	} else if (file_sectors == zlo->zone_capacity) {
		zone->cond = BLK_ZONE_COND_FULL;
		zone->wp = ULLONG_MAX;
	} else {
		zone->cond = BLK_ZONE_COND_CLOSED;
		zone->wp = zone->start + file_sectors;
	}
	spin_unlock_irqrestore(&zone->wp_lock, flags);

	return 0;
}

static int zloop_open_zone(struct zloop_device *zlo, unsigned int zone_no)
{
	struct zloop_zone *zone = &zlo->zones[zone_no];
	int ret = 0;

	if (test_bit(ZLOOP_ZONE_CONV, &zone->flags))
		return -EIO;

	mutex_lock(&zone->lock);

	if (test_and_clear_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags)) {
		ret = zloop_update_seq_zone(zlo, zone_no);
		if (ret)
			goto unlock;
	}

	switch (zone->cond) {
	case BLK_ZONE_COND_EXP_OPEN:
		break;
	case BLK_ZONE_COND_EMPTY:
	case BLK_ZONE_COND_CLOSED:
	case BLK_ZONE_COND_IMP_OPEN:
		zone->cond = BLK_ZONE_COND_EXP_OPEN;
		break;
	case BLK_ZONE_COND_FULL:
	default:
		ret = -EIO;
		break;
	}

unlock:
	mutex_unlock(&zone->lock);

	return ret;
}

static int zloop_close_zone(struct zloop_device *zlo, unsigned int zone_no)
{
	struct zloop_zone *zone = &zlo->zones[zone_no];
	unsigned long flags;
	int ret = 0;

	if (test_bit(ZLOOP_ZONE_CONV, &zone->flags))
		return -EIO;

	mutex_lock(&zone->lock);

	if (test_and_clear_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags)) {
		ret = zloop_update_seq_zone(zlo, zone_no);
		if (ret)
			goto unlock;
	}

	switch (zone->cond) {
	case BLK_ZONE_COND_CLOSED:
		break;
	case BLK_ZONE_COND_IMP_OPEN:
	case BLK_ZONE_COND_EXP_OPEN:
		spin_lock_irqsave(&zone->wp_lock, flags);
		if (zone->wp == zone->start)
			zone->cond = BLK_ZONE_COND_EMPTY;
		else
			zone->cond = BLK_ZONE_COND_CLOSED;
		spin_unlock_irqrestore(&zone->wp_lock, flags);
		break;
	case BLK_ZONE_COND_EMPTY:
	case BLK_ZONE_COND_FULL:
	default:
		ret = -EIO;
		break;
	}

unlock:
	mutex_unlock(&zone->lock);

	return ret;
}

static int zloop_reset_zone(struct zloop_device *zlo, unsigned int zone_no)
{
	struct zloop_zone *zone = &zlo->zones[zone_no];
	unsigned long flags;
	int ret = 0;

	if (test_bit(ZLOOP_ZONE_CONV, &zone->flags))
		return -EIO;

	mutex_lock(&zone->lock);

	if (!test_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags) &&
	    zone->cond == BLK_ZONE_COND_EMPTY)
		goto unlock;

	if (vfs_truncate(&zone->file->f_path, 0)) {
		set_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags);
		ret = -EIO;
		goto unlock;
	}

	spin_lock_irqsave(&zone->wp_lock, flags);
	zone->cond = BLK_ZONE_COND_EMPTY;
	zone->wp = zone->start;
	clear_bit(ZLOOP_ZONE_SEQ_ERROR, &zone->flags);
	spin_unlock_irqrestore(&zone->wp_lock, flags);

unlock:
	mutex_unlock(&zone->lock);

	return ret;
}

static int zloop_reset_all_zones(struct zloop_device *zlo)
{
	unsigned int i;
	int ret;

	for (i = zlo->nr_conv_zones; i < zlo->nr_zones; i++) {
		ret = zloop_reset_zone(zlo, i);
		if (ret)
			return ret;
	}

	return 0;
}

static int zloop_finish_zone(struct zloop_device *zlo, unsigned int zone_no)
{
	struct zloop_zone *zone = &zlo->zones[zone_no];
	unsigned long flags;
	int ret = 0;

	if (test_bit(ZLOOP_ZONE_CONV, &zone-&g