[thirdparty/linux.git] / block / blk-zoned.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Zoned block device handling
 *
 * Copyright (c) 2015, Hannes Reinecke
 * Copyright (c) 2015, SUSE Linux GmbH
 *
 * Copyright (c) 2016, Damien Le Moal
 * Copyright (c) 2016, Western Digital
 */

#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/rbtree.h>
#include <linux/blkdev.h>
#include <linux/blk-mq.h>
#include <linux/mm.h>
#include <linux/vmalloc.h>
#include <linux/sched/mm.h>

#include "blk.h"

#define ZONE_COND_NAME(name) [BLK_ZONE_COND_##name] = #name
static const char *const zone_cond_name[] = {
	ZONE_COND_NAME(NOT_WP),
	ZONE_COND_NAME(EMPTY),
	ZONE_COND_NAME(IMP_OPEN),
	ZONE_COND_NAME(EXP_OPEN),
	ZONE_COND_NAME(CLOSED),
	ZONE_COND_NAME(READONLY),
	ZONE_COND_NAME(FULL),
	ZONE_COND_NAME(OFFLINE),
};
#undef ZONE_COND_NAME

/**
 * blk_zone_cond_str - Return string XXX in BLK_ZONE_COND_XXX.
 * @zone_cond: BLK_ZONE_COND_XXX.
 *
 * Description: Centralize block layer function to convert BLK_ZONE_COND_XXX
 * into string format. Useful in the debugging and tracing zone conditions. For
 * invalid BLK_ZONE_COND_XXX it returns string "UNKNOWN".
 */
const char *blk_zone_cond_str(enum blk_zone_cond zone_cond)
{
	static const char *zone_cond_str = "UNKNOWN";

	if (zone_cond < ARRAY_SIZE(zone_cond_name) && zone_cond_name[zone_cond])
		zone_cond_str = zone_cond_name[zone_cond];

	return zone_cond_str;
}
EXPORT_SYMBOL_GPL(blk_zone_cond_str);

static inline sector_t blk_zone_start(struct request_queue *q,
				      sector_t sector)
{
	sector_t zone_mask = blk_queue_zone_sectors(q) - 1;

	return sector & ~zone_mask;
}

/*
 * Return true if a request is a write requests that needs zone write locking.
 */
bool blk_req_needs_zone_write_lock(struct request *rq)
{
	if (!rq->q->seq_zones_wlock)
		return false;

	if (blk_rq_is_passthrough(rq))
		return false;

	switch (req_op(rq)) {
	case REQ_OP_WRITE_ZEROES:
	case REQ_OP_WRITE_SAME:
	case REQ_OP_WRITE:
		return blk_rq_zone_is_seq(rq);
	default:
		return false;
	}
}
EXPORT_SYMBOL_GPL(blk_req_needs_zone_write_lock);

void __blk_req_zone_write_lock(struct request *rq)
{
	if (WARN_ON_ONCE(test_and_set_bit(blk_rq_zone_no(rq),
					  rq->q->seq_zones_wlock)))
		return;

	WARN_ON_ONCE(rq->rq_flags & RQF_ZONE_WRITE_LOCKED);
	rq->rq_flags |= RQF_ZONE_WRITE_LOCKED;
}
EXPORT_SYMBOL_GPL(__blk_req_zone_write_lock);

void __blk_req_zone_write_unlock(struct request *rq)
{
	rq->rq_flags &= ~RQF_ZONE_WRITE_LOCKED;
	if (rq->q->seq_zones_wlock)
		WARN_ON_ONCE(!test_and_clear_bit(blk_rq_zone_no(rq),
						 rq->q->seq_zones_wlock));
}
EXPORT_SYMBOL_GPL(__blk_req_zone_write_unlock);

/**
 * blkdev_nr_zones - Get number of zones
 * @disk:	Target gendisk
 *
 * Return the total number of zones of a zoned block device.  For a block
 * device without zone capabilities, the number of zones is always 0.
 */
unsigned int blkdev_nr_zones(struct gendisk *disk)
{
	sector_t zone_sectors = blk_queue_zone_sectors(disk->queue);

	if (!blk_queue_is_zoned(disk->queue))
		return 0;
	return (get_capacity(disk) + zone_sectors - 1) >> ilog2(zone_sectors);
}
EXPORT_SYMBOL_GPL(blkdev_nr_zones);

/**
 * blkdev_report_zones - Get zones information
 * @bdev:	Target block device
 * @sector:	Sector from which to report zones
 * @nr_zones:	Maximum number of zones to report
 * @cb:		Callback function called for each reported zone
 * @data:	Private data for the callback
 *
 * Description:
 *    Get zone information starting from the zone containing @sector for at most
 *    @nr_zones, and call @cb for each zone reported by the device.
 *    To report all zones in a device starting from @sector, the BLK_ALL_ZONES
 *    constant can be passed to @nr_zones.
 *    Returns the number of zones reported by the device, or a negative errno
 *    value in case of failure.
 *
 *    Note: The caller must use memalloc_noXX_save/restore() calls to control
 *    memory allocations done within this function.
 */
int blkdev_report_zones(struct block_device *bdev, sector_t sector,
			unsigned int nr_zones, report_zones_cb cb, void *data)
{
	struct gendisk *disk = bdev->bd_disk;
	sector_t capacity = get_capacity(disk);

	if (!blk_queue_is_zoned(bdev_get_queue(bdev)) ||
	    WARN_ON_ONCE(!disk->fops->report_zones))
		return -EOPNOTSUPP;

	if (!nr_zones || sector >= capacity)
		return 0;

	return disk->fops->report_zones(disk, sector, nr_zones, cb, data);
}
EXPORT_SYMBOL_GPL(blkdev_report_zones);

static inline bool blkdev_allow_reset_all_zones(struct block_device *bdev,
						sector_t sector,
						sector_t nr_sectors)
{
	if (!blk_queue_zone_resetall(bdev_get_queue(bdev)))
		return false;

	/*
	 * REQ_OP_ZONE_RESET_ALL can be executed only if the number of sectors
	 * of the applicable zone range is the entire disk.
	 */
	return !sector && nr_sectors == get_capacity(bdev->bd_disk);
}

/**
 * blkdev_zone_mgmt - Execute a zone management operation on a range of zones
 * @bdev:	Target block device
 * @op:		Operation to be performed on the zones
 * @sector:	Start sector of the first zone to operate on
 * @nr_sectors:	Number of sectors, should be at least the length of one zone and
 *		must be zone size aligned.
 * @gfp_mask:	Memory allocation flags (for bio_alloc)
 *
 * Description:
 *    Perform the specified operation on the range of zones specified by
 *    @sector..@sector+@nr_sectors. Specifying the entire disk sector range
 *    is valid, but the specified range should not contain conventional zones.
 *    The operation to execute on each zone can be a zone reset, open, close
 *    or finish request.
 */
int blkdev_zone_mgmt(struct block_device *bdev, enum req_opf op,
		     sector_t sector, sector_t nr_sectors,
		     gfp_t gfp_mask)
{
	struct request_queue *q = bdev_get_queue(bdev);
	sector_t zone_sectors = blk_queue_zone_sectors(q);
	sector_t capacity = get_capacity(bdev->bd_disk);
	sector_t end_sector = sector + nr_sectors;
	struct bio *bio = NULL;
	int ret;

	if (!blk_queue_is_zoned(q))
		return -EOPNOTSUPP;

	if (bdev_read_only(bdev))
		return -EPERM;

	if (!op_is_zone_mgmt(op))
		return -EOPNOTSUPP;

	if (end_sector <= sector || end_sector > capacity)
		/* Out of range */
		return -EINVAL;

	/* Check alignment (handle eventual smaller last zone) */
	if (sector & (zone_sectors - 1))
		return -EINVAL;

	if ((nr_sectors & (zone_sectors - 1)) && end_sector != capacity)
		return -EINVAL;

	while (sector < end_sector) {
		bio = blk_next_bio(bio, 0, gfp_mask);
		bio_set_dev(bio, bdev);

		/*
		 * Special case for the zone reset operation that reset all
		 * zones, this is useful for applications like mkfs.
		 */
		if (op == REQ_OP_ZONE_RESET &&
		    blkdev_allow_reset_all_zones(bdev, sector, nr_sectors)) {
			bio->bi_opf = REQ_OP_ZONE_RESET_ALL;
			break;
		}

		bio->bi_opf = op | REQ_SYNC;
		bio->bi_iter.bi_sector = sector;
		sector += zone_sectors;

		/* This may take a while, so be nice to others */
		cond_resched();
	}

	ret = submit_bio_wait(bio);
	bio_put(bio);

	return ret;
}
EXPORT_SYMBOL_GPL(blkdev_zone_mgmt);

struct zone_report_args {
	struct blk_zone __user *zones;
};

static int blkdev_copy_zone_to_user(struct blk_zone *zone, unsigned int idx,
				    void *data)
{
	struct zone_report_args *args = data;

	if (copy_to_user(&args->zones[idx], zone, sizeof(struct blk_zone)))
		return -EFAULT;
	return 0;
}

/*
 * BLKREPORTZONE ioctl processing.
 * Called from blkdev_ioctl.
 */
int blkdev_report_zones_ioctl(struct block_device *bdev, fmode_t mode,
			      unsigned int cmd, unsigned long arg)
{
	void __user *argp = (void __user *)arg;
	struct zone_report_args args;
	struct request_queue *q;
	struct blk_zone_report rep;
	int ret;

	if (!argp)
		return -EINVAL;

	q = bdev_get_queue(bdev);
	if (!q)
		return -ENXIO;

	if (!blk_queue_is_zoned(q))
		return -ENOTTY;

	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;

	if (copy_from_user(&rep, argp, sizeof(struct blk_zone_report)))
		return -EFAULT;

	if (!rep.nr_zones)
		return -EINVAL;

	args.zones = argp + sizeof(struct blk_zone_report);
	ret = blkdev_report_zones(bdev, rep.sector, rep.nr_zones,
				  blkdev_copy_zone_to_user, &args);
	if (ret < 0)
		return ret;

	rep.nr_zones = ret;
	if (copy_to_user(argp, &rep, sizeof(struct blk_zone_report)))
		return -EFAULT;
	return 0;
}

/*
 * BLKRESETZONE, BLKOPENZONE, BLKCLOSEZONE and BLKFINISHZONE ioctl processing.
 * Called from blkdev_ioctl.
 */
int blkdev_zone_mgmt_ioctl(struct block_device *bdev, fmode_t mode,
			   unsigned int cmd, unsigned long arg)
{
	void __user *argp = (void __user *)arg;
	struct request_queue *q;
	struct blk_zone_range zrange;
	enum req_opf op;

	if (!argp)
		return -EINVAL;

	q = bdev_get_queue(bdev);
	if (!q)
		return -ENXIO;

	if (!blk_queue_is_zoned(q))
		return -ENOTTY;

	if (!capable(CAP_SYS_ADMIN))
		return -EACCES;

	if (!(mode & FMODE_WRITE))
		return -EBADF;

	if (copy_from_user(&zrange, argp, sizeof(struct blk_zone_range)))
		return -EFAULT;

	switch (cmd) {
	case BLKRESETZONE:
		op = REQ_OP_ZONE_RESET;
		break;
	case BLKOPENZONE:
		op = REQ_OP_ZONE_OPEN;
		break;
	case BLKCLOSEZONE:
		op = REQ_OP_ZONE_CLOSE;
		break;
	case BLKFINISHZONE:
		op = REQ_OP_ZONE_FINISH;
		break;
	default:
		return -ENOTTY;
	}

	return blkdev_zone_mgmt(bdev, op, zrange.sector, zrange.nr_sectors,
				GFP_KERNEL);
}

static inline unsigned long *blk_alloc_zone_bitmap(int node,
						   unsigned int nr_zones)
{
	return kcalloc_node(BITS_TO_LONGS(nr_zones), sizeof(unsigned long),
			    GFP_NOIO, node);
}

void blk_queue_free_zone_bitmaps(struct request_queue *q)
{
	kfree(q->conv_zones_bitmap);
	q->conv_zones_bitmap = NULL;
	kfree(q->seq_zones_wlock);
	q->seq_zones_wlock = NULL;
}

struct blk_revalidate_zone_args {
	struct gendisk	*disk;
	unsigned long	*conv_zones_bitmap;
	unsigned long	*seq_zones_wlock;
	unsigned int	nr_zones;
	sector_t	zone_sectors;
	sector_t	sector;
};

/*
 * Helper function to check the validity of zones of a zoned block device.
 */
static int blk_revalidate_zone_cb(struct blk_zone *zone, unsigned int idx,
				  void *data)
{
	struct blk_revalidate_zone_args *args = data;
	struct gendisk *disk = args->disk;
	struct request_queue *q = disk->queue;
	sector_t capacity = get_capacity(disk);

	/*
	 * All zones must have the same size, with the exception on an eventual
	 * smaller last zone.
	 */
	if (zone->start == 0) {
		if (zone->len == 0 || !is_power_of_2(zone->len)) {
			pr_warn("%s: Invalid zoned device with non power of two zone size (%llu)\n",
				disk->disk_name, zone->len);
			return -ENODEV;
		}

		args->zone_sectors = zone->len;
		args->nr_zones = (capacity + zone->len - 1) >> ilog2(zone->len);
	} else if (zone->start + args->zone_sectors < capacity) {
		if (zone->len != args->zone_sectors) {
			pr_warn("%s: Invalid zoned device with non constant zone size\n",
				disk->disk_name);
			return -ENODEV;
		}
	} else {
		if (zone->len > args->zone_sectors) {
			pr_warn("%s: Invalid zoned device with larger last zone size\n",
				disk->disk_name);
			return -ENODEV;
		}
	}

	/* Check for holes in the zone report */
	if (zone->start != args->sector) {
		pr_warn("%s: Zone gap at sectors %llu..%llu\n",
			disk->disk_name, args->sector, zone->start);
		return -ENODEV;
	}

	/* Check zone type */
	switch (zone->type) {
	case BLK_ZONE_TYPE_CONVENTIONAL:
		if (!args->conv_zones_bitmap) {
			args->conv_zones_bitmap =
				blk_alloc_zone_bitmap(q->node, args->nr_zones);
			if (!args->conv_zones_bitmap)
				return -ENOMEM;
		}
		set_bit(idx, args->conv_zones_bitmap);
		break;
	case BLK_ZONE_TYPE_SEQWRITE_REQ:
	case BLK_ZONE_TYPE_SEQWRITE_PREF:
		if (!args->seq_zones_wlock) {
			args->seq_zones_wlock =
				blk_alloc_zone_bitmap(q->node, args->nr_zones);
			if (!args->seq_zones_wlock)
				return -ENOMEM;
		}
		break;
	default:
		pr_warn("%s: Invalid zone type 0x%x at sectors %llu\n",
			disk->disk_name, (int)zone->type, zone->start);
		return -ENODEV;
	}

	args->sector += zone->len;
	return 0;
}

/**
 * blk_revalidate_disk_zones - (re)allocate and initialize zone bitmaps
 * @disk:	Target disk
 *
 * Helper function for low-level device drivers to (re) allocate and initialize
 * a disk request queue zone bitmaps. This functions should normally be called
 * within the disk ->revalidate method for blk-mq based drivers.  For BIO based
 * drivers only q->nr_zones needs to be updated so that the sysfs exposed value
 * is correct.
 */
int blk_revalidate_disk_zones(struct gendisk *disk)
{
	struct request_queue *q = disk->queue;
	struct blk_revalidate_zone_args args = {
		.disk		= disk,
	};
	unsigned int noio_flag;
	int ret;

	if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
		return -EIO;
	if (WARN_ON_ONCE(!queue_is_mq(q)))
		return -EIO;

	/*
	 * Ensure that all memory allocations in this context are done as if
	 * GFP_NOIO was specified.
	 */
	noio_flag = memalloc_noio_save();
	ret = disk->fops->report_zones(disk, 0, UINT_MAX,
				       blk_revalidate_zone_cb, &args);
	memalloc_noio_restore(noio_flag);

	/*
	 * Install the new bitmaps and update nr_zones only once the queue is
	 * stopped and all I/Os are completed (i.e. a scheduler is not
	 * referencing the bitmaps).
	 */
	blk_mq_freeze_queue(q);
	if (ret >= 0) {
		blk_queue_chunk_sectors(q, args.zone_sectors);
		q->nr_zones = args.nr_zones;
		swap(q->seq_zones_wlock, args.seq_zones_wlock);
		swap(q->conv_zones_bitmap, args.conv_zones_bitmap);
		ret = 0;
	} else {
		pr_warn("%s: failed to revalidate zones\n", disk->disk_name);
		blk_queue_free_zone_bitmaps(q);
	}
	blk_mq_unfreeze_queue(q);

	kfree(args.seq_zones_wlock);
	kfree(args.conv_zones_bitmap);
	return ret;
}
EXPORT_SYMBOL_GPL(blk_revalidate_disk_zones);
Commit	Line	Data
3dcf60bc	1	// SPDX-License-Identifier: GPL-2.0
6a0cb1bc HR	2	/*
	3	* Zoned block device handling
	4	*
	5	* Copyright (c) 2015, Hannes Reinecke
	6	* Copyright (c) 2015, SUSE Linux GmbH
	7	*
	8	* Copyright (c) 2016, Damien Le Moal
	9	* Copyright (c) 2016, Western Digital
	10	*/
	11
	12	#include <linux/kernel.h>
	13	#include <linux/module.h>
	14	#include <linux/rbtree.h>
	15	#include <linux/blkdev.h>
bf505456	16	#include <linux/blk-mq.h>
26202928 DLM	17	#include <linux/mm.h>
26202928 DLM	18	#include <linux/vmalloc.h>
bd976e52	19	#include <linux/sched/mm.h>
6a0cb1bc	20
a2d6b3a2 DLM	21	#include "blk.h"
a2d6b3a2 DLM	22
02694e86 CK	23	#define ZONE_COND_NAME(name) [BLK_ZONE_COND_##name] = #name
	24	static const char *const zone_cond_name[] = {
	25	ZONE_COND_NAME(NOT_WP),
	26	ZONE_COND_NAME(EMPTY),
	27	ZONE_COND_NAME(IMP_OPEN),
	28	ZONE_COND_NAME(EXP_OPEN),
	29	ZONE_COND_NAME(CLOSED),
	30	ZONE_COND_NAME(READONLY),
	31	ZONE_COND_NAME(FULL),
	32	ZONE_COND_NAME(OFFLINE),
	33	};
	34	#undef ZONE_COND_NAME
	35
	36	/**
	37	* blk_zone_cond_str - Return string XXX in BLK_ZONE_COND_XXX.
	38	* @zone_cond: BLK_ZONE_COND_XXX.
	39	*
	40	* Description: Centralize block layer function to convert BLK_ZONE_COND_XXX
	41	* into string format. Useful in the debugging and tracing zone conditions. For
	42	* invalid BLK_ZONE_COND_XXX it returns string "UNKNOWN".
	43	*/
	44	const char *blk_zone_cond_str(enum blk_zone_cond zone_cond)
	45	{
	46	static const char *zone_cond_str = "UNKNOWN";
	47
	48	if (zone_cond < ARRAY_SIZE(zone_cond_name) && zone_cond_name[zone_cond])
	49	zone_cond_str = zone_cond_name[zone_cond];
	50
	51	return zone_cond_str;
	52	}
	53	EXPORT_SYMBOL_GPL(blk_zone_cond_str);
	54
6a0cb1bc HR	55	static inline sector_t blk_zone_start(struct request_queue *q,
	56	sector_t sector)
	57	{
f99e8648	58	sector_t zone_mask = blk_queue_zone_sectors(q) - 1;
6a0cb1bc HR	59
	60	return sector & ~zone_mask;
	61	}
	62
6cc77e9c CH	63	/*
	64	* Return true if a request is a write requests that needs zone write locking.
	65	*/
	66	bool blk_req_needs_zone_write_lock(struct request *rq)
	67	{
	68	if (!rq->q->seq_zones_wlock)
	69	return false;
	70
	71	if (blk_rq_is_passthrough(rq))
	72	return false;
	73
	74	switch (req_op(rq)) {
	75	case REQ_OP_WRITE_ZEROES:
	76	case REQ_OP_WRITE_SAME:
	77	case REQ_OP_WRITE:
	78	return blk_rq_zone_is_seq(rq);
	79	default:
	80	return false;
	81	}
	82	}
	83	EXPORT_SYMBOL_GPL(blk_req_needs_zone_write_lock);
	84
	85	void __blk_req_zone_write_lock(struct request *rq)
	86	{
	87	if (WARN_ON_ONCE(test_and_set_bit(blk_rq_zone_no(rq),
	88	rq->q->seq_zones_wlock)))
	89	return;
	90
	91	WARN_ON_ONCE(rq->rq_flags & RQF_ZONE_WRITE_LOCKED);
	92	rq->rq_flags \|= RQF_ZONE_WRITE_LOCKED;
	93	}
	94	EXPORT_SYMBOL_GPL(__blk_req_zone_write_lock);
	95
	96	void __blk_req_zone_write_unlock(struct request *rq)
	97	{
	98	rq->rq_flags &= ~RQF_ZONE_WRITE_LOCKED;
	99	if (rq->q->seq_zones_wlock)
	100	WARN_ON_ONCE(!test_and_clear_bit(blk_rq_zone_no(rq),
	101	rq->q->seq_zones_wlock));
	102	}
	103	EXPORT_SYMBOL_GPL(__blk_req_zone_write_unlock);
	104
a91e1380 DLM	105	/**
a91e1380 DLM	106	* blkdev_nr_zones - Get number of zones
9b38bb4b	107	* @disk: Target gendisk
a91e1380	108	*
9b38bb4b CH	109	* Return the total number of zones of a zoned block device. For a block
9b38bb4b CH	110	* device without zone capabilities, the number of zones is always 0.
a91e1380	111	*/
9b38bb4b	112	unsigned int blkdev_nr_zones(struct gendisk *disk)
a91e1380	113	{
9b38bb4b	114	sector_t zone_sectors = blk_queue_zone_sectors(disk->queue);
a91e1380	115
9b38bb4b	116	if (!blk_queue_is_zoned(disk->queue))
a91e1380	117	return 0;
9b38bb4b	118	return (get_capacity(disk) + zone_sectors - 1) >> ilog2(zone_sectors);
a91e1380 DLM	119	}
	120	EXPORT_SYMBOL_GPL(blkdev_nr_zones);
	121
6a0cb1bc HR	122	/**
	123	* blkdev_report_zones - Get zones information
	124	* @bdev: Target block device
	125	* @sector: Sector from which to report zones
d4100351 CH	126	* @nr_zones: Maximum number of zones to report
	127	* @cb: Callback function called for each reported zone
	128	* @data: Private data for the callback
6a0cb1bc HR	129	*
6a0cb1bc HR	130	* Description:
d4100351 CH	131	* Get zone information starting from the zone containing @sector for at most
	132	* @nr_zones, and call @cb for each zone reported by the device.
	133	* To report all zones in a device starting from @sector, the BLK_ALL_ZONES
	134	* constant can be passed to @nr_zones.
	135	* Returns the number of zones reported by the device, or a negative errno
	136	* value in case of failure.
	137	*
	138	* Note: The caller must use memalloc_noXX_save/restore() calls to control
	139	* memory allocations done within this function.
6a0cb1bc	140	*/
e76239a3	141	int blkdev_report_zones(struct block_device *bdev, sector_t sector,
d4100351	142	unsigned int nr_zones, report_zones_cb cb, void *data)
6a0cb1bc	143	{
ceeb373a	144	struct gendisk *disk = bdev->bd_disk;
5eac3eb3	145	sector_t capacity = get_capacity(disk);
6a0cb1bc	146
d4100351 CH	147	if (!blk_queue_is_zoned(bdev_get_queue(bdev)) \|\|
d4100351 CH	148	WARN_ON_ONCE(!disk->fops->report_zones))
e76239a3	149	return -EOPNOTSUPP;
6a0cb1bc	150
d4100351	151	if (!nr_zones \|\| sector >= capacity)
6a0cb1bc	152	return 0;
6a0cb1bc	153
d4100351	154	return disk->fops->report_zones(disk, sector, nr_zones, cb, data);
6a0cb1bc HR	155	}
	156	EXPORT_SYMBOL_GPL(blkdev_report_zones);
	157
6e33dbf2	158	static inline bool blkdev_allow_reset_all_zones(struct block_device *bdev,
c7a1d926	159	sector_t sector,
6e33dbf2 CK	160	sector_t nr_sectors)
	161	{
	162	if (!blk_queue_zone_resetall(bdev_get_queue(bdev)))
	163	return false;
	164
6e33dbf2	165	/*
5eac3eb3 DLM	166	* REQ_OP_ZONE_RESET_ALL can be executed only if the number of sectors
5eac3eb3 DLM	167	* of the applicable zone range is the entire disk.
6e33dbf2	168	*/
5eac3eb3	169	return !sector && nr_sectors == get_capacity(bdev->bd_disk);
6e33dbf2 CK	170	}
6e33dbf2 CK	171
6a0cb1bc	172	/**
6c1b1da5	173	* blkdev_zone_mgmt - Execute a zone management operation on a range of zones
6a0cb1bc	174	* @bdev: Target block device
6c1b1da5 AJ	175	* @op: Operation to be performed on the zones
	176	* @sector: Start sector of the first zone to operate on
	177	* @nr_sectors: Number of sectors, should be at least the length of one zone and
	178	* must be zone size aligned.
6a0cb1bc HR	179	* @gfp_mask: Memory allocation flags (for bio_alloc)
	180	*
	181	* Description:
6c1b1da5	182	* Perform the specified operation on the range of zones specified by
6a0cb1bc HR	183	* @sector..@sector+@nr_sectors. Specifying the entire disk sector range
6a0cb1bc HR	184	* is valid, but the specified range should not contain conventional zones.
6c1b1da5 AJ	185	* The operation to execute on each zone can be a zone reset, open, close
6c1b1da5 AJ	186	* or finish request.
6a0cb1bc	187	*/
6c1b1da5 AJ	188	int blkdev_zone_mgmt(struct block_device *bdev, enum req_opf op,
	189	sector_t sector, sector_t nr_sectors,
	190	gfp_t gfp_mask)
6a0cb1bc HR	191	{
6a0cb1bc HR	192	struct request_queue *q = bdev_get_queue(bdev);
6c1b1da5	193	sector_t zone_sectors = blk_queue_zone_sectors(q);
5eac3eb3	194	sector_t capacity = get_capacity(bdev->bd_disk);
6a0cb1bc	195	sector_t end_sector = sector + nr_sectors;
a2d6b3a2	196	struct bio *bio = NULL;
6a0cb1bc HR	197	int ret;
6a0cb1bc HR	198
6a0cb1bc HR	199	if (!blk_queue_is_zoned(q))
	200	return -EOPNOTSUPP;
	201
a2d6b3a2 DLM	202	if (bdev_read_only(bdev))
	203	return -EPERM;
	204
6c1b1da5 AJ	205	if (!op_is_zone_mgmt(op))
	206	return -EOPNOTSUPP;
	207
11bde986	208	if (end_sector <= sector \|\| end_sector > capacity)
6a0cb1bc HR	209	/* Out of range */
	210	return -EINVAL;
	211
	212	/* Check alignment (handle eventual smaller last zone) */
6a0cb1bc HR	213	if (sector & (zone_sectors - 1))
	214	return -EINVAL;
	215
5eac3eb3	216	if ((nr_sectors & (zone_sectors - 1)) && end_sector != capacity)
6a0cb1bc HR	217	return -EINVAL;
	218
	219	while (sector < end_sector) {
a2d6b3a2	220	bio = blk_next_bio(bio, 0, gfp_mask);
74d46992	221	bio_set_dev(bio, bdev);
6a0cb1bc	222
c7a1d926 DLM	223	/*
	224	* Special case for the zone reset operation that reset all
	225	* zones, this is useful for applications like mkfs.
	226	*/
6c1b1da5 AJ	227	if (op == REQ_OP_ZONE_RESET &&
6c1b1da5 AJ	228	blkdev_allow_reset_all_zones(bdev, sector, nr_sectors)) {
c7a1d926 DLM	229	bio->bi_opf = REQ_OP_ZONE_RESET_ALL;
	230	break;
	231	}
	232
8e42d239	233	bio->bi_opf = op \| REQ_SYNC;
c7a1d926	234	bio->bi_iter.bi_sector = sector;
6a0cb1bc HR	235	sector += zone_sectors;
	236
	237	/* This may take a while, so be nice to others */
	238	cond_resched();
6a0cb1bc HR	239	}
6a0cb1bc HR	240
a2d6b3a2 DLM	241	ret = submit_bio_wait(bio);
	242	bio_put(bio);
	243
a2d6b3a2	244	return ret;
6a0cb1bc	245	}
6c1b1da5	246	EXPORT_SYMBOL_GPL(blkdev_zone_mgmt);
3ed05a98	247
d4100351 CH	248	struct zone_report_args {
	249	struct blk_zone __user *zones;
	250	};
	251
	252	static int blkdev_copy_zone_to_user(struct blk_zone *zone, unsigned int idx,
	253	void *data)
	254	{
	255	struct zone_report_args *args = data;
	256
	257	if (copy_to_user(&args->zones[idx], zone, sizeof(struct blk_zone)))
	258	return -EFAULT;
	259	return 0;
	260	}
	261
56c4bddb	262	/*
3ed05a98 ST	263	* BLKREPORTZONE ioctl processing.
	264	* Called from blkdev_ioctl.
	265	*/
	266	int blkdev_report_zones_ioctl(struct block_device *bdev, fmode_t mode,
	267	unsigned int cmd, unsigned long arg)
	268	{
	269	void __user argp = (void __user )arg;
d4100351	270	struct zone_report_args args;
3ed05a98 ST	271	struct request_queue *q;
3ed05a98 ST	272	struct blk_zone_report rep;
3ed05a98 ST	273	int ret;
	274
	275	if (!argp)
	276	return -EINVAL;
	277
	278	q = bdev_get_queue(bdev);
	279	if (!q)
	280	return -ENXIO;
	281
	282	if (!blk_queue_is_zoned(q))
	283	return -ENOTTY;
	284
	285	if (!capable(CAP_SYS_ADMIN))
	286	return -EACCES;
	287
	288	if (copy_from_user(&rep, argp, sizeof(struct blk_zone_report)))
	289	return -EFAULT;
	290
	291	if (!rep.nr_zones)
	292	return -EINVAL;
	293
d4100351 CH	294	args.zones = argp + sizeof(struct blk_zone_report);
	295	ret = blkdev_report_zones(bdev, rep.sector, rep.nr_zones,
	296	blkdev_copy_zone_to_user, &args);
	297	if (ret < 0)
	298	return ret;
3ed05a98	299
d4100351 CH	300	rep.nr_zones = ret;
	301	if (copy_to_user(argp, &rep, sizeof(struct blk_zone_report)))
	302	return -EFAULT;
	303	return 0;
3ed05a98 ST	304	}
3ed05a98 ST	305
56c4bddb	306	/*
e876df1f	307	* BLKRESETZONE, BLKOPENZONE, BLKCLOSEZONE and BLKFINISHZONE ioctl processing.
3ed05a98 ST	308	* Called from blkdev_ioctl.
3ed05a98 ST	309	*/
e876df1f AJ	310	int blkdev_zone_mgmt_ioctl(struct block_device *bdev, fmode_t mode,
e876df1f AJ	311	unsigned int cmd, unsigned long arg)
3ed05a98 ST	312	{
	313	void __user argp = (void __user )arg;
	314	struct request_queue *q;
	315	struct blk_zone_range zrange;
e876df1f	316	enum req_opf op;
3ed05a98 ST	317
	318	if (!argp)
	319	return -EINVAL;
	320
	321	q = bdev_get_queue(bdev);
	322	if (!q)
	323	return -ENXIO;
	324
	325	if (!blk_queue_is_zoned(q))
	326	return -ENOTTY;
	327
	328	if (!capable(CAP_SYS_ADMIN))
	329	return -EACCES;
	330
	331	if (!(mode & FMODE_WRITE))
	332	return -EBADF;
	333
	334	if (copy_from_user(&zrange, argp, sizeof(struct blk_zone_range)))
	335	return -EFAULT;
	336
e876df1f AJ	337	switch (cmd) {
	338	case BLKRESETZONE:
	339	op = REQ_OP_ZONE_RESET;
	340	break;
	341	case BLKOPENZONE:
	342	op = REQ_OP_ZONE_OPEN;
	343	break;
	344	case BLKCLOSEZONE:
	345	op = REQ_OP_ZONE_CLOSE;
	346	break;
	347	case BLKFINISHZONE:
	348	op = REQ_OP_ZONE_FINISH;
	349	break;
	350	default:
	351	return -ENOTTY;
	352	}
	353
	354	return blkdev_zone_mgmt(bdev, op, zrange.sector, zrange.nr_sectors,
	355	GFP_KERNEL);
3ed05a98	356	}
bf505456 DLM	357
	358	static inline unsigned long *blk_alloc_zone_bitmap(int node,
	359	unsigned int nr_zones)
	360	{
	361	return kcalloc_node(BITS_TO_LONGS(nr_zones), sizeof(unsigned long),
	362	GFP_NOIO, node);
	363	}
	364
bf505456 DLM	365	void blk_queue_free_zone_bitmaps(struct request_queue *q)
bf505456 DLM	366	{
f216fdd7 CH	367	kfree(q->conv_zones_bitmap);
f216fdd7 CH	368	q->conv_zones_bitmap = NULL;
bf505456 DLM	369	kfree(q->seq_zones_wlock);
	370	q->seq_zones_wlock = NULL;
	371	}
	372
d4100351 CH	373	struct blk_revalidate_zone_args {
d4100351 CH	374	struct gendisk *disk;
f216fdd7	375	unsigned long *conv_zones_bitmap;
d4100351	376	unsigned long *seq_zones_wlock;
e94f5819	377	unsigned int nr_zones;
6c6b3549	378	sector_t zone_sectors;
d4100351 CH	379	sector_t sector;
	380	};
	381
d9dd7308 DLM	382	/*
	383	* Helper function to check the validity of zones of a zoned block device.
	384	*/
d4100351 CH	385	static int blk_revalidate_zone_cb(struct blk_zone *zone, unsigned int idx,
d4100351 CH	386	void *data)
d9dd7308	387	{
d4100351 CH	388	struct blk_revalidate_zone_args *args = data;
d4100351 CH	389	struct gendisk *disk = args->disk;
d9dd7308	390	struct request_queue *q = disk->queue;
d9dd7308 DLM	391	sector_t capacity = get_capacity(disk);
	392
	393	/*
	394	* All zones must have the same size, with the exception on an eventual
	395	* smaller last zone.
	396	*/
6c6b3549 CH	397	if (zone->start == 0) {
	398	if (zone->len == 0 \|\| !is_power_of_2(zone->len)) {
	399	pr_warn("%s: Invalid zoned device with non power of two zone size (%llu)\n",
	400	disk->disk_name, zone->len);
	401	return -ENODEV;
	402	}
d9dd7308	403
6c6b3549 CH	404	args->zone_sectors = zone->len;
	405	args->nr_zones = (capacity + zone->len - 1) >> ilog2(zone->len);
	406	} else if (zone->start + args->zone_sectors < capacity) {
	407	if (zone->len != args->zone_sectors) {
	408	pr_warn("%s: Invalid zoned device with non constant zone size\n",
	409	disk->disk_name);
	410	return -ENODEV;
	411	}
	412	} else {
	413	if (zone->len > args->zone_sectors) {
	414	pr_warn("%s: Invalid zoned device with larger last zone size\n",
	415	disk->disk_name);
	416	return -ENODEV;
	417	}
d9dd7308 DLM	418	}
	419
	420	/* Check for holes in the zone report */
d4100351	421	if (zone->start != args->sector) {
d9dd7308	422	pr_warn("%s: Zone gap at sectors %llu..%llu\n",
d4100351 CH	423	disk->disk_name, args->sector, zone->start);
d4100351 CH	424	return -ENODEV;
d9dd7308 DLM	425	}
	426
	427	/* Check zone type */
	428	switch (zone->type) {
	429	case BLK_ZONE_TYPE_CONVENTIONAL:
e94f5819 CH	430	if (!args->conv_zones_bitmap) {
	431	args->conv_zones_bitmap =
	432	blk_alloc_zone_bitmap(q->node, args->nr_zones);
	433	if (!args->conv_zones_bitmap)
	434	return -ENOMEM;
	435	}
	436	set_bit(idx, args->conv_zones_bitmap);
	437	break;
d9dd7308 DLM	438	case BLK_ZONE_TYPE_SEQWRITE_REQ:
d9dd7308 DLM	439	case BLK_ZONE_TYPE_SEQWRITE_PREF:
e94f5819 CH	440	if (!args->seq_zones_wlock) {
	441	args->seq_zones_wlock =
	442	blk_alloc_zone_bitmap(q->node, args->nr_zones);
	443	if (!args->seq_zones_wlock)
	444	return -ENOMEM;
	445	}
d9dd7308 DLM	446	break;
	447	default:
	448	pr_warn("%s: Invalid zone type 0x%x at sectors %llu\n",
	449	disk->disk_name, (int)zone->type, zone->start);
d4100351	450	return -ENODEV;
d9dd7308 DLM	451	}
d9dd7308 DLM	452
d4100351 CH	453	args->sector += zone->len;
	454	return 0;
	455	}
	456
bf505456 DLM	457	/**
	458	* blk_revalidate_disk_zones - (re)allocate and initialize zone bitmaps
	459	* @disk: Target disk
	460	*
	461	* Helper function for low-level device drivers to (re) allocate and initialize
	462	* a disk request queue zone bitmaps. This functions should normally be called
ae58954d CH	463	* within the disk ->revalidate method for blk-mq based drivers. For BIO based
	464	* drivers only q->nr_zones needs to be updated so that the sysfs exposed value
	465	* is correct.
bf505456 DLM	466	*/
	467	int blk_revalidate_disk_zones(struct gendisk *disk)
	468	{
	469	struct request_queue *q = disk->queue;
e94f5819 CH	470	struct blk_revalidate_zone_args args = {
e94f5819 CH	471	.disk = disk,
e94f5819	472	};
6c6b3549 CH	473	unsigned int noio_flag;
6c6b3549 CH	474	int ret;
bf505456	475
c98c3d09 CH	476	if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
c98c3d09 CH	477	return -EIO;
ae58954d CH	478	if (WARN_ON_ONCE(!queue_is_mq(q)))
ae58954d CH	479	return -EIO;
bf505456	480
e94f5819	481	/*
6c6b3549 CH	482	* Ensure that all memory allocations in this context are done as if
6c6b3549 CH	483	* GFP_NOIO was specified.
e94f5819	484	*/
6c6b3549 CH	485	noio_flag = memalloc_noio_save();
	486	ret = disk->fops->report_zones(disk, 0, UINT_MAX,
	487	blk_revalidate_zone_cb, &args);
	488	memalloc_noio_restore(noio_flag);
bf505456	489
bf505456	490	/*
6c6b3549 CH	491	* Install the new bitmaps and update nr_zones only once the queue is
	492	* stopped and all I/Os are completed (i.e. a scheduler is not
	493	* referencing the bitmaps).
bf505456 DLM	494	*/
bf505456 DLM	495	blk_mq_freeze_queue(q);
d4100351	496	if (ret >= 0) {
6c6b3549	497	blk_queue_chunk_sectors(q, args.zone_sectors);
e94f5819	498	q->nr_zones = args.nr_zones;
d4100351	499	swap(q->seq_zones_wlock, args.seq_zones_wlock);
f216fdd7	500	swap(q->conv_zones_bitmap, args.conv_zones_bitmap);
d4100351 CH	501	ret = 0;
d4100351 CH	502	} else {
bf505456	503	pr_warn("%s: failed to revalidate zones\n", disk->disk_name);
bf505456	504	blk_queue_free_zone_bitmaps(q);
bf505456	505	}
d4100351	506	blk_mq_unfreeze_queue(q);
bf505456	507
d4100351	508	kfree(args.seq_zones_wlock);
f216fdd7	509	kfree(args.conv_zones_bitmap);
bf505456 DLM	510	return ret;
	511	}
	512	EXPORT_SYMBOL_GPL(blk_revalidate_disk_zones);