Commit 72421690 authored by Jens Axboe's avatar Jens Axboe
Browse files

Merge tag 'md-6.17-20250819' of... 

Merge tag 'md-6.17-20250819' of gitolite.kernel.org:pub/scm/linux/kernel/git/mdraid/linux into block-6.17

Pull MD fixes from Yu:

"- Add a legacy_async_del_gendisk mode, to prevent a user tools
   regression. New user tools releases will not use such a mode, the old
   release with a new kernel now will have warning about deprecated
   behavior, and we prepare to remove this legacy mode after about a
   year later.
 - The rename in kernel causing user tools build failure, revert the
   rename in mdp_superblock_s.
 - Fix a regression that interrupted resync can be shown as recover from
   mdstat or sysfs."

* tag 'md-6.17-20250819' of gitolite.kernel.org:pub/scm/linux/kernel/git/mdraid/linux:
  md: fix sync_action incorrect display during resync
  md: add helper rdev_needs_recovery()
  md: keep recovery_cp in mdp_superblock_s
  md: add legacy_async_del_gendisk mode
parents f4ae1744 b7ee30f0

drivers/md/md.c

+92 −30
Original line number Diff line number Diff line
@@ -339,6 +339,7 @@ static int start_readonly; 
 * so all the races disappear.

 */

static bool create_on_open = true;

static bool legacy_async_del_gendisk = true;



/*

 * We have a system wide 'event count' that is incremented
@@ -877,7 +878,9 @@ void mddev_unlock(struct mddev *mddev) 
		export_rdev(rdev, mddev);

	}



	/* Call del_gendisk after release reconfig_mutex to avoid

	if (!legacy_async_del_gendisk) {

		/*

		 * Call del_gendisk after release reconfig_mutex to avoid

		 * deadlock (e.g. call del_gendisk under the lock and an

		 * access to sysfs files waits the lock)

		 * And MD_DELETED is only used for md raid which is set in
@@ -887,6 +890,7 @@ void mddev_unlock(struct mddev *mddev) 
		if (test_bit(MD_DELETED, &mddev->flags))

			del_gendisk(mddev->gendisk);

	}

}

EXPORT_SYMBOL_GPL(mddev_unlock);



struct md_rdev *md_find_rdev_nr_rcu(struct mddev *mddev, int nr)
@@ -1419,7 +1423,7 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *freshest, stru 
		else {

			if (sb->events_hi == sb->cp_events_hi &&

				sb->events_lo == sb->cp_events_lo) {

				mddev->resync_offset = sb->resync_offset;

				mddev->resync_offset = sb->recovery_cp;

			} else

				mddev->resync_offset = 0;

		}
@@ -1547,13 +1551,13 @@ static void super_90_sync(struct mddev *mddev, struct md_rdev *rdev) 
	mddev->minor_version = sb->minor_version;

	if (mddev->in_sync)

	{

		sb->resync_offset = mddev->resync_offset;

		sb->recovery_cp = mddev->resync_offset;

		sb->cp_events_hi = (mddev->events>>32);

		sb->cp_events_lo = (u32)mddev->events;

		if (mddev->resync_offset == MaxSector)

			sb->state = (1<< MD_SB_CLEAN);

	} else

		sb->resync_offset = 0;

		sb->recovery_cp = 0;



	sb->layout = mddev->layout;

	sb->chunk_size = mddev->chunk_sectors << 9;
@@ -4835,9 +4839,42 @@ metadata_store(struct mddev *mddev, const char *buf, size_t len) 
static struct md_sysfs_entry md_metadata =

__ATTR_PREALLOC(metadata_version, S_IRUGO|S_IWUSR, metadata_show, metadata_store);



static bool rdev_needs_recovery(struct md_rdev *rdev, sector_t sectors)

{

	return rdev->raid_disk >= 0 &&

	       !test_bit(Journal, &rdev->flags) &&

	       !test_bit(Faulty, &rdev->flags) &&

	       !test_bit(In_sync, &rdev->flags) &&

	       rdev->recovery_offset < sectors;

}



static enum sync_action md_get_active_sync_action(struct mddev *mddev)

{

	struct md_rdev *rdev;

	bool is_recover = false;



	if (mddev->resync_offset < MaxSector)

		return ACTION_RESYNC;



	if (mddev->reshape_position != MaxSector)

		return ACTION_RESHAPE;



	rcu_read_lock();

	rdev_for_each_rcu(rdev, mddev) {

		if (rdev_needs_recovery(rdev, MaxSector)) {

			is_recover = true;

			break;

		}

	}

	rcu_read_unlock();



	return is_recover ? ACTION_RECOVER : ACTION_IDLE;

}



enum sync_action md_sync_action(struct mddev *mddev)

{

	unsigned long recovery = mddev->recovery;

	enum sync_action active_action;



	/*

	 * frozen has the highest priority, means running sync_thread will be
@@ -4861,8 +4898,17 @@ enum sync_action md_sync_action(struct mddev *mddev) 
	    !test_bit(MD_RECOVERY_NEEDED, &recovery))

		return ACTION_IDLE;



	if (test_bit(MD_RECOVERY_RESHAPE, &recovery) ||

	    mddev->reshape_position != MaxSector)

	/*

	 * Check if any sync operation (resync/recover/reshape) is

	 * currently active. This ensures that only one sync operation

	 * can run at a time. Returns the type of active operation, or

	 * ACTION_IDLE if none are active.

	 */

	active_action = md_get_active_sync_action(mddev);

	if (active_action != ACTION_IDLE)

		return active_action;



	if (test_bit(MD_RECOVERY_RESHAPE, &recovery))

		return ACTION_RESHAPE;



	if (test_bit(MD_RECOVERY_RECOVER, &recovery))
@@ -5818,6 +5864,13 @@ static void md_kobj_release(struct kobject *ko) 
{

	struct mddev *mddev = container_of(ko, struct mddev, kobj);



	if (legacy_async_del_gendisk) {

		if (mddev->sysfs_state)

			sysfs_put(mddev->sysfs_state);

		if (mddev->sysfs_level)

			sysfs_put(mddev->sysfs_level);

		del_gendisk(mddev->gendisk);

	}

	put_disk(mddev->gendisk);

}
@@ -6021,6 +6074,9 @@ static int md_alloc_and_put(dev_t dev, char *name) 
{

	struct mddev *mddev = md_alloc(dev, name);



	if (legacy_async_del_gendisk)

		pr_warn("md: async del_gendisk mode will be removed in future, please upgrade to mdadm-4.5+\n");



	if (IS_ERR(mddev))

		return PTR_ERR(mddev);

	mddev_put(mddev);
@@ -6431,10 +6487,22 @@ static void md_clean(struct mddev *mddev) 
	mddev->persistent = 0;

	mddev->level = LEVEL_NONE;

	mddev->clevel[0] = 0;



	/*

	 * For legacy_async_del_gendisk mode, it can stop the array in the

	 * middle of assembling it, then it still can access the array. So

	 * it needs to clear MD_CLOSING. If not legacy_async_del_gendisk,

	 * it can't open the array again after stopping it. So it doesn't

	 * clear MD_CLOSING.

	 */

	if (legacy_async_del_gendisk && mddev->hold_active) {

		clear_bit(MD_CLOSING, &mddev->flags);

	} else {

		/* if UNTIL_STOP is set, it's cleared here */

		mddev->hold_active = 0;

		/* Don't clear MD_CLOSING, or mddev can be opened again. */

		mddev->flags &= BIT_ULL_MASK(MD_CLOSING);

	}

	mddev->sb_flags = 0;

	mddev->ro = MD_RDWR;

	mddev->metadata_type[0] = 0;
@@ -6658,6 +6726,7 @@ static int do_md_stop(struct mddev *mddev, int mode) 


		export_array(mddev);

		md_clean(mddev);

		if (!legacy_async_del_gendisk)

			set_bit(MD_DELETED, &mddev->flags);

	}

	md_new_event();
@@ -8968,11 +9037,7 @@ static sector_t md_sync_position(struct mddev *mddev, enum sync_action action) 
		start = MaxSector;

		rcu_read_lock();

		rdev_for_each_rcu(rdev, mddev)

			if (rdev->raid_disk >= 0 &&

			    !test_bit(Journal, &rdev->flags) &&

			    !test_bit(Faulty, &rdev->flags) &&

			    !test_bit(In_sync, &rdev->flags) &&

			    rdev->recovery_offset < start)

			if (rdev_needs_recovery(rdev, start))

				start = rdev->recovery_offset;

		rcu_read_unlock();
@@ -9331,12 +9396,8 @@ void md_do_sync(struct md_thread *thread) 
			    test_bit(MD_RECOVERY_RECOVER, &mddev->recovery)) {

				rcu_read_lock();

				rdev_for_each_rcu(rdev, mddev)

					if (rdev->raid_disk >= 0 &&

					    mddev->delta_disks >= 0 &&

					    !test_bit(Journal, &rdev->flags) &&

					    !test_bit(Faulty, &rdev->flags) &&

					    !test_bit(In_sync, &rdev->flags) &&

					    rdev->recovery_offset < mddev->curr_resync)

					if (mddev->delta_disks >= 0 &&

					    rdev_needs_recovery(rdev, mddev->curr_resync))

						rdev->recovery_offset = mddev->curr_resync;

				rcu_read_unlock();

			}
@@ -10392,6 +10453,7 @@ module_param_call(start_ro, set_ro, get_ro, NULL, S_IRUSR|S_IWUSR); 
module_param(start_dirty_degraded, int, S_IRUGO|S_IWUSR);

module_param_call(new_array, add_named_array, NULL, NULL, S_IWUSR);

module_param(create_on_open, bool, S_IRUSR|S_IWUSR);

module_param(legacy_async_del_gendisk, bool, 0600);



MODULE_LICENSE("GPL");

MODULE_DESCRIPTION("MD RAID framework");

include/uapi/linux/raid/md_p.h

+1 −1
Original line number Diff line number Diff line
@@ -173,7 +173,7 @@ typedef struct mdp_superblock_s { 
#else

#error unspecified endianness

#endif

	__u32 resync_offset;	/* 11 resync checkpoint sector count	      */

	__u32 recovery_cp;	/* 11 resync checkpoint sector count	      */

	/* There are only valid for minor_version > 90 */

	__u64 reshape_position;	/* 12,13 next address in array-space for reshape */

	__u32 new_level;	/* 14 new level we are reshaping to	      */