Merge tag 'for-6.11/block-20240722' of git://git.kernel.dk/linux

Pull more block updates from Jens Axboe:

 - MD fixes via Song:
     - md-cluster fixes (Heming Zhao)
     - raid1 fix (Mateusz Jończyk)

 - s390/dasd module description (Jeff)

 - Series cleaning up and hardening the blk-mq debugfs flag handling
   (John, Christoph)

 - blk-cgroup cleanup (Xiu)

 - Error polled IO attempts if backend doesn't support it (hexue)

 - Fix for an sbitmap hang (Yang)

* tag 'for-6.11/block-20240722' of git://git.kernel.dk/linux: (23 commits)
  blk-cgroup: move congestion_count to struct blkcg
  sbitmap: fix io hung due to race on sbitmap_word::cleared
  block: avoid polling configuration errors
  block: Catch possible entries missing from rqf_name[]
  block: Simplify definition of RQF_NAME()
  block: Use enum to define RQF_x bit indexes
  block: Catch possible entries missing from cmd_flag_name[]
  block: Catch possible entries missing from alloc_policy_name[]
  block: Catch possible entries missing from hctx_flag_name[]
  block: Catch possible entries missing from hctx_state_name[]
  block: Catch possible entries missing from blk_queue_flag_name[]
  block: Make QUEUE_FLAG_x as an enum
  block: Relocate BLK_MQ_MAX_DEPTH
  block: Relocate BLK_MQ_CPU_WORK_BATCH
  block: remove QUEUE_FLAG_STOPPED
  block: Add missing entry to hctx_flag_name[]
  block: Add zone write plugging entry to rqf_name[]
  block: Add missing entries from cmd_flag_name[]
  s390/dasd: fix error checks in dasd_copy_pair_store()
  s390/dasd: add missing MODULE_DESCRIPTION() macros
  ...

block/blk-cgroup.c

@@ -2182,12 +2182,13 @@ void blk_cgroup_bio_start(struct bio *bio)
 
 bool blk_cgroup_congested(void)
 {
-	struct cgroup_subsys_state *css;
+	struct blkcg *blkcg;
 	bool ret = false;
 
 	rcu_read_lock();
-	for (css = blkcg_css(); css; css = css->parent) {
-		if (atomic_read(&css->cgroup->congestion_count)) {
+	for (blkcg = css_to_blkcg(blkcg_css()); blkcg;
+	     blkcg = blkcg_parent(blkcg)) {
+		if (atomic_read(&blkcg->congestion_count)) {
 			ret = true;
 			break;
 		}

block/blk-cgroup.h

@@ -95,6 +95,8 @@ struct blkcg {
 	struct cgroup_subsys_state	css;
 	spinlock_t			lock;
 	refcount_t			online_pin;
+	/* If there is block congestion on this cgroup. */
+	atomic_t			congestion_count;
 
 	struct radix_tree_root		blkg_tree;
 	struct blkcg_gq	__rcu		*blkg_hint;
@@ -374,7 +376,7 @@ static inline void blkcg_use_delay(struct blkcg_gq *blkg)
 	if (WARN_ON_ONCE(atomic_read(&blkg->use_delay) < 0))
 		return;
 	if (atomic_add_return(1, &blkg->use_delay) == 1)
-		atomic_inc(&blkg->blkcg->css.cgroup->congestion_count);
+		atomic_inc(&blkg->blkcg->congestion_count);
 }
 
 static inline int blkcg_unuse_delay(struct blkcg_gq *blkg)
@@ -399,7 +401,7 @@ static inline int blkcg_unuse_delay(struct blkcg_gq *blkg)
 	if (old == 0)
 		return 0;
 	if (old == 1)
-		atomic_dec(&blkg->blkcg->css.cgroup->congestion_count);
+		atomic_dec(&blkg->blkcg->congestion_count);
 	return 1;
 }
 
@@ -418,7 +420,7 @@ static inline void blkcg_set_delay(struct blkcg_gq *blkg, u64 delay)
 
 	/* We only want 1 person setting the congestion count for this blkg. */
 	if (!old && atomic_try_cmpxchg(&blkg->use_delay, &old, -1))
-		atomic_inc(&blkg->blkcg->css.cgroup->congestion_count);
+		atomic_inc(&blkg->blkcg->congestion_count);
 
 	atomic64_set(&blkg->delay_nsec, delay);
 }
@@ -435,7 +437,7 @@ static inline void blkcg_clear_delay(struct blkcg_gq *blkg)
 
 	/* We only want 1 person clearing the congestion count for this blkg. */
 	if (old && atomic_try_cmpxchg(&blkg->use_delay, &old, 0))
-		atomic_dec(&blkg->blkcg->css.cgroup->congestion_count);
+		atomic_dec(&blkg->blkcg->congestion_count);
 }
 
 /**

block/blk-core.c

@@ -791,8 +791,11 @@ void submit_bio_noacct(struct bio *bio)
 		}
 	}
 
-	if (!(q->limits.features & BLK_FEAT_POLL))
+	if (!(q->limits.features & BLK_FEAT_POLL) &&
+			(bio->bi_opf & REQ_POLLED)) {
 		bio_clear_polled(bio);
+		goto not_supported;
+	}
 
 	switch (bio_op(bio)) {
 	case REQ_OP_READ:

block/blk-mq-debugfs.c

@@ -5,6 +5,7 @@
 
 #include <linux/kernel.h>
 #include <linux/blkdev.h>
+#include <linux/build_bug.h>
 #include <linux/debugfs.h>
 
 #include "blk.h"
@@ -79,7 +80,6 @@ static int queue_pm_only_show(void *data, struct seq_file *m)
 
 #define QUEUE_FLAG_NAME(name) [QUEUE_FLAG_##name] = #name
 static const char *const blk_queue_flag_name[] = {
-	QUEUE_FLAG_NAME(STOPPED),
 	QUEUE_FLAG_NAME(DYING),
 	QUEUE_FLAG_NAME(NOMERGES),
 	QUEUE_FLAG_NAME(SAME_COMP),
@@ -100,6 +100,7 @@ static int queue_state_show(void *data, struct seq_file *m)
 {
 	struct request_queue *q = data;
 
+	BUILD_BUG_ON(ARRAY_SIZE(blk_queue_flag_name) != QUEUE_FLAG_MAX);
 	blk_flags_show(m, q->queue_flags, blk_queue_flag_name,
 		       ARRAY_SIZE(blk_queue_flag_name));
 	seq_puts(m, "\n");
@@ -164,6 +165,7 @@ static int hctx_state_show(void *data, struct seq_file *m)
 {
 	struct blk_mq_hw_ctx *hctx = data;
 
+	BUILD_BUG_ON(ARRAY_SIZE(hctx_state_name) != BLK_MQ_S_MAX);
 	blk_flags_show(m, hctx->state, hctx_state_name,
 		       ARRAY_SIZE(hctx_state_name));
 	seq_puts(m, "\n");
@@ -181,10 +183,11 @@ static const char *const alloc_policy_name[] = {
 static const char *const hctx_flag_name[] = {
 	HCTX_FLAG_NAME(SHOULD_MERGE),
 	HCTX_FLAG_NAME(TAG_QUEUE_SHARED),
-	HCTX_FLAG_NAME(BLOCKING),
-	HCTX_FLAG_NAME(NO_SCHED),
 	HCTX_FLAG_NAME(STACKING),
 	HCTX_FLAG_NAME(TAG_HCTX_SHARED),
+	HCTX_FLAG_NAME(BLOCKING),
+	HCTX_FLAG_NAME(NO_SCHED),
+	HCTX_FLAG_NAME(NO_SCHED_BY_DEFAULT),
 };
 #undef HCTX_FLAG_NAME
 
@@ -193,6 +196,10 @@ static int hctx_flags_show(void *data, struct seq_file *m)
 	struct blk_mq_hw_ctx *hctx = data;
 	const int alloc_policy = BLK_MQ_FLAG_TO_ALLOC_POLICY(hctx->flags);
 
+	BUILD_BUG_ON(ARRAY_SIZE(hctx_flag_name) !=
+			BLK_MQ_F_ALLOC_POLICY_START_BIT);
+	BUILD_BUG_ON(ARRAY_SIZE(alloc_policy_name) != BLK_TAG_ALLOC_MAX);
+
 	seq_puts(m, "alloc_policy=");
 	if (alloc_policy < ARRAY_SIZE(alloc_policy_name) &&
 	    alloc_policy_name[alloc_policy])
@@ -223,12 +230,17 @@ static const char *const cmd_flag_name[] = {
 	CMD_FLAG_NAME(RAHEAD),
 	CMD_FLAG_NAME(BACKGROUND),
 	CMD_FLAG_NAME(NOWAIT),
-	CMD_FLAG_NAME(NOUNMAP),
 	CMD_FLAG_NAME(POLLED),
+	CMD_FLAG_NAME(ALLOC_CACHE),
+	CMD_FLAG_NAME(SWAP),
+	CMD_FLAG_NAME(DRV),
+	CMD_FLAG_NAME(FS_PRIVATE),
+	CMD_FLAG_NAME(ATOMIC),
+	CMD_FLAG_NAME(NOUNMAP),
 };
 #undef CMD_FLAG_NAME
 
-#define RQF_NAME(name) [ilog2((__force u32)RQF_##name)] = #name
+#define RQF_NAME(name) [__RQF_##name] = #name
 static const char *const rqf_name[] = {
 	RQF_NAME(STARTED),
 	RQF_NAME(FLUSH_SEQ),
@@ -243,6 +255,7 @@ static const char *const rqf_name[] = {
 	RQF_NAME(HASHED),
 	RQF_NAME(STATS),
 	RQF_NAME(SPECIAL_PAYLOAD),
+	RQF_NAME(ZONE_WRITE_PLUGGING),
 	RQF_NAME(TIMED_OUT),
 	RQF_NAME(RESV),
 };
@@ -268,6 +281,9 @@ int __blk_mq_debugfs_rq_show(struct seq_file *m, struct request *rq)
 	const enum req_op op = req_op(rq);
 	const char *op_str = blk_op_str(op);
 
+	BUILD_BUG_ON(ARRAY_SIZE(cmd_flag_name) != __REQ_NR_BITS);
+	BUILD_BUG_ON(ARRAY_SIZE(rqf_name) != __RQF_BITS);
+
 	seq_printf(m, "%p {.op=", rq);
 	if (strcmp(op_str, "UNKNOWN") == 0)
 		seq_printf(m, "%u", op);

block/blk-mq.h

@@ -36,6 +36,8 @@ enum {
 	BLK_MQ_TAG_MAX		= BLK_MQ_NO_TAG - 1,
 };
 
+#define BLK_MQ_CPU_WORK_BATCH	(8)
+
 typedef unsigned int __bitwise blk_insert_t;
 #define BLK_MQ_INSERT_AT_HEAD		((__force blk_insert_t)0x01)
 

drivers/md/md-cluster.c

@@ -15,6 +15,7 @@
 
 #define LVB_SIZE	64
 #define NEW_DEV_TIMEOUT 5000
+#define WAIT_DLM_LOCK_TIMEOUT (30 * HZ)
 
 struct dlm_lock_resource {
 	dlm_lockspace_t *ls;
@@ -56,6 +57,7 @@ struct resync_info {
 #define		MD_CLUSTER_ALREADY_IN_CLUSTER		6
 #define		MD_CLUSTER_PENDING_RECV_EVENT		7
 #define 	MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD		8
+#define		MD_CLUSTER_WAITING_FOR_SYNC		9
 
 struct md_cluster_info {
 	struct mddev *mddev; /* the md device which md_cluster_info belongs to */
@@ -91,6 +93,7 @@ struct md_cluster_info {
 	sector_t sync_hi;
 };
 
+/* For compatibility, add the new msg_type at the end. */
 enum msg_type {
 	METADATA_UPDATED = 0,
 	RESYNCING,
@@ -100,6 +103,7 @@ enum msg_type {
 	BITMAP_NEEDS_SYNC,
 	CHANGE_CAPACITY,
 	BITMAP_RESIZE,
+	RESYNCING_START,
 };
 
 struct cluster_msg {
@@ -130,8 +134,13 @@ static int dlm_lock_sync(struct dlm_lock_resource *res, int mode)
 			0, sync_ast, res, res->bast);
 	if (ret)
 		return ret;
-	wait_event(res->sync_locking, res->sync_locking_done);
+	ret = wait_event_timeout(res->sync_locking, res->sync_locking_done,
+				WAIT_DLM_LOCK_TIMEOUT);
 	res->sync_locking_done = false;
+	if (!ret) {
+		pr_err("locking DLM '%s' timeout!\n", res->name);
+		return -EBUSY;
+	}
 	if (res->lksb.sb_status == 0)
 		res->mode = mode;
 	return res->lksb.sb_status;
@@ -455,6 +464,7 @@ static void process_suspend_info(struct mddev *mddev,
 		clear_bit(MD_RESYNCING_REMOTE, &mddev->recovery);
 		remove_suspend_info(mddev, slot);
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
+		clear_bit(MD_CLUSTER_WAITING_FOR_SYNC, &cinfo->state);
 		md_wakeup_thread(mddev->thread);
 		return;
 	}
@@ -525,6 +535,7 @@ static int process_add_new_disk(struct mddev *mddev, struct cluster_msg *cmsg)
 		res = -1;
 	}
 	clear_bit(MD_CLUSTER_WAITING_FOR_NEWDISK, &cinfo->state);
+	set_bit(MD_CLUSTER_WAITING_FOR_SYNC, &cinfo->state);
 	return res;
 }
 
@@ -593,6 +604,9 @@ static int process_recvd_msg(struct mddev *mddev, struct cluster_msg *msg)
 	case CHANGE_CAPACITY:
 		set_capacity_and_notify(mddev->gendisk, mddev->array_sectors);
 		break;
+	case RESYNCING_START:
+		clear_bit(MD_CLUSTER_WAITING_FOR_SYNC, &mddev->cluster_info->state);
+		break;
 	case RESYNCING:
 		set_bit(MD_RESYNCING_REMOTE, &mddev->recovery);
 		process_suspend_info(mddev, le32_to_cpu(msg->slot),
@@ -743,7 +757,7 @@ static void unlock_comm(struct md_cluster_info *cinfo)
  */
 static int __sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg)
 {
-	int error;
+	int error, unlock_error;
 	int slot = cinfo->slot_number - 1;
 
 	cmsg->slot = cpu_to_le32(slot);
@@ -751,7 +765,7 @@ static int __sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg)
 	error = dlm_lock_sync(cinfo->message_lockres, DLM_LOCK_EX);
 	if (error) {
 		pr_err("md-cluster: failed to get EX on MESSAGE (%d)\n", error);
-		goto failed_message;
+		return error;
 	}
 
 	memcpy(cinfo->message_lockres->lksb.sb_lvbptr, (void *)cmsg,
@@ -781,14 +795,10 @@ static int __sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg)
 	}
 
 failed_ack:
-	error = dlm_unlock_sync(cinfo->message_lockres);
-	if (unlikely(error != 0)) {
+	while ((unlock_error = dlm_unlock_sync(cinfo->message_lockres)))
 		pr_err("md-cluster: failed convert to NL on MESSAGE(%d)\n",
-			error);
-		/* in case the message can't be released due to some reason */
-		goto failed_ack;
-	}
-failed_message:
+			unlock_error);
+
 	return error;
 }
 
@@ -1343,6 +1353,23 @@ static void resync_info_get(struct mddev *mddev, sector_t *lo, sector_t *hi)
 	spin_unlock_irq(&cinfo->suspend_lock);
 }
 
+static int resync_status_get(struct mddev *mddev)
+{
+	struct md_cluster_info *cinfo = mddev->cluster_info;
+
+	return test_bit(MD_CLUSTER_WAITING_FOR_SYNC, &cinfo->state);
+}
+
+static int resync_start_notify(struct mddev *mddev)
+{
+	struct md_cluster_info *cinfo = mddev->cluster_info;
+	struct cluster_msg cmsg = {0};
+
+	cmsg.type = cpu_to_le32(RESYNCING_START);
+
+	return sendmsg(cinfo, &cmsg, 0);
+}
+
 static int resync_info_update(struct mddev *mddev, sector_t lo, sector_t hi)
 {
 	struct md_cluster_info *cinfo = mddev->cluster_info;
@@ -1577,6 +1604,8 @@ static const struct md_cluster_operations cluster_ops = {
 	.resync_start = resync_start,
 	.resync_finish = resync_finish,
 	.resync_info_update = resync_info_update,
+	.resync_start_notify = resync_start_notify,
+	.resync_status_get = resync_status_get,
 	.resync_info_get = resync_info_get,
 	.metadata_update_start = metadata_update_start,
 	.metadata_update_finish = metadata_update_finish,

drivers/md/md-cluster.h

@@ -14,6 +14,8 @@ struct md_cluster_operations {
 	int (*leave)(struct mddev *mddev);
 	int (*slot_number)(struct mddev *mddev);
 	int (*resync_info_update)(struct mddev *mddev, sector_t lo, sector_t hi);
+	int (*resync_start_notify)(struct mddev *mddev);
+	int (*resync_status_get)(struct mddev *mddev);
 	void (*resync_info_get)(struct mddev *mddev, sector_t *lo, sector_t *hi);
 	int (*metadata_update_start)(struct mddev *mddev);
 	int (*metadata_update_finish)(struct mddev *mddev);

drivers/md/md.c

@@ -8978,7 +8978,8 @@ void md_do_sync(struct md_thread *thread)
 	 * This will mean we have to start checking from the beginning again.
 	 *
 	 */
-
+	if (mddev_is_clustered(mddev))
+		md_cluster_ops->resync_start_notify(mddev);
 	do {
 		int mddev2_minor = -1;
 		mddev->curr_resync = MD_RESYNC_DELAYED;
@@ -9992,8 +9993,18 @@ static void check_sb_changes(struct mddev *mddev, struct md_rdev *rdev)
 			 */
 			if (rdev2->raid_disk == -1 && role != MD_DISK_ROLE_SPARE &&
 			    !(le32_to_cpu(sb->feature_map) &
-			      MD_FEATURE_RESHAPE_ACTIVE)) {
-				rdev2->saved_raid_disk = role;
+			      MD_FEATURE_RESHAPE_ACTIVE) &&
+			    !md_cluster_ops->resync_status_get(mddev)) {
+				/*
+				 * -1 to make raid1_add_disk() set conf->fullsync
+				 * to 1. This could avoid skipping sync when the
+				 * remote node is down during resyncing.
+				 */
+				if ((le32_to_cpu(sb->feature_map)
+				    & MD_FEATURE_RECOVERY_OFFSET))
+					rdev2->saved_raid_disk = -1;
+				else
+					rdev2->saved_raid_disk = role;
 				ret = remove_and_add_spares(mddev, rdev2);
 				pr_info("Activated spare: %pg\n",
 					rdev2->bdev);

drivers/md/raid1.c

@@ -680,6 +680,7 @@ static int choose_slow_rdev(struct r1conf *conf, struct r1bio *r1_bio,
 		len = r1_bio->sectors;
 		read_len = raid1_check_read_range(rdev, this_sector, &len);
 		if (read_len == r1_bio->sectors) {
+			*max_sectors = read_len;
 			update_read_sectors(conf, disk, this_sector, read_len);
 			return disk;
 		}

drivers/s390/block/dasd_devmap.c

@@ -2248,13 +2248,19 @@ static ssize_t dasd_copy_pair_store(struct device *dev,
 
 	/* allocate primary devmap if needed */
 	prim_devmap = dasd_find_busid(prim_busid);
-	if (IS_ERR(prim_devmap))
+	if (IS_ERR(prim_devmap)) {
 		prim_devmap = dasd_add_busid(prim_busid, DASD_FEATURE_DEFAULT);
+		if (IS_ERR(prim_devmap))
+			return PTR_ERR(prim_devmap);
+	}
 
 	/* allocate secondary devmap if needed */
 	sec_devmap = dasd_find_busid(sec_busid);
-	if (IS_ERR(sec_devmap))
+	if (IS_ERR(sec_devmap)) {
 		sec_devmap = dasd_add_busid(sec_busid, DASD_FEATURE_DEFAULT);
+		if (IS_ERR(sec_devmap))
+			return PTR_ERR(sec_devmap);
+	}
 
 	/* setting copy relation is only allowed for offline secondary */
 	if (sec_devmap->device)

drivers/s390/block/dasd_diag.c

@@ -29,6 +29,7 @@
 #include "dasd_int.h"
 #include "dasd_diag.h"
 
+MODULE_DESCRIPTION("S/390 Support for DIAG access to DASD Disks");
 MODULE_LICENSE("GPL");
 
 /* The maximum number of blocks per request (max_blocks) is dependent on the

drivers/s390/block/dasd_eckd.c

@@ -44,6 +44,7 @@
 /* 64k are 128 x 512 byte sectors  */
 #define DASD_RAW_SECTORS_PER_TRACK 128
 
+MODULE_DESCRIPTION("S/390 DASD ECKD Disks device driver");
 MODULE_LICENSE("GPL");
 
 static struct dasd_discipline dasd_eckd_discipline;

drivers/s390/block/dasd_fba.c

@@ -32,6 +32,7 @@
 #define DASD_FBA_CCW_LOCATE 0x43
 #define DASD_FBA_CCW_DEFINE_EXTENT 0x63
 
+MODULE_DESCRIPTION("S/390 DASD FBA Disks device driver");
 MODULE_LICENSE("GPL");
 
 static struct dasd_discipline dasd_fba_discipline;

include/linux/blk-mq.h

@@ -27,38 +27,61 @@ typedef enum rq_end_io_ret (rq_end_io_fn)(struct request *, blk_status_t);
  * request flags */
 typedef __u32 __bitwise req_flags_t;
 
-/* drive already may have started this one */
-#define RQF_STARTED		((__force req_flags_t)(1 << 1))
-/* request for flush sequence */
-#define RQF_FLUSH_SEQ		((__force req_flags_t)(1 << 4))
-/* merge of different types, fail separately */
-#define RQF_MIXED_MERGE		((__force req_flags_t)(1 << 5))
-/* don't call prep for this one */
-#define RQF_DONTPREP		((__force req_flags_t)(1 << 7))
-/* use hctx->sched_tags */
-#define RQF_SCHED_TAGS		((__force req_flags_t)(1 << 8))
-/* use an I/O scheduler for this request */
-#define RQF_USE_SCHED		((__force req_flags_t)(1 << 9))
-/* vaguely specified driver internal error.  Ignored by the block layer */
-#define RQF_FAILED		((__force req_flags_t)(1 << 10))
-/* don't warn about errors */
-#define RQF_QUIET		((__force req_flags_t)(1 << 11))
-/* account into disk and partition IO statistics */
-#define RQF_IO_STAT		((__force req_flags_t)(1 << 13))
-/* runtime pm request */
-#define RQF_PM			((__force req_flags_t)(1 << 15))
-/* on IO scheduler merge hash */
-#define RQF_HASHED		((__force req_flags_t)(1 << 16))
-/* track IO completion time */
-#define RQF_STATS		((__force req_flags_t)(1 << 17))
-/* Look at ->special_vec for the actual data payload instead of the
-   bio chain. */
-#define RQF_SPECIAL_PAYLOAD	((__force req_flags_t)(1 << 18))
-/* The request completion needs to be signaled to zone write pluging. */
-#define RQF_ZONE_WRITE_PLUGGING	((__force req_flags_t)(1 << 20))
-/* ->timeout has been called, don't expire again */
-#define RQF_TIMED_OUT		((__force req_flags_t)(1 << 21))
-#define RQF_RESV		((__force req_flags_t)(1 << 23))
+/* Keep rqf_name[] in sync with the definitions below */
+enum {
+	/* drive already may have started this one */
+	__RQF_STARTED,
+	/* request for flush sequence */
+	__RQF_FLUSH_SEQ,
+	/* merge of different types, fail separately */
+	__RQF_MIXED_MERGE,
+	/* don't call prep for this one */
+	__RQF_DONTPREP,
+	/* use hctx->sched_tags */
+	__RQF_SCHED_TAGS,
+	/* use an I/O scheduler for this request */
+	__RQF_USE_SCHED,
+	/* vaguely specified driver internal error.  Ignored by block layer */
+	__RQF_FAILED,
+	/* don't warn about errors */
+	__RQF_QUIET,
+	/* account into disk and partition IO statistics */
+	__RQF_IO_STAT,
+	/* runtime pm request */
+	__RQF_PM,
+	/* on IO scheduler merge hash */
+	__RQF_HASHED,
+	/* track IO completion time */
+	__RQF_STATS,
+	/* Look at ->special_vec for the actual data payload instead of the
+	   bio chain. */
+	__RQF_SPECIAL_PAYLOAD,
+	/* request completion needs to be signaled to zone write plugging. */
+	__RQF_ZONE_WRITE_PLUGGING,
+	/* ->timeout has been called, don't expire again */
+	__RQF_TIMED_OUT,
+	__RQF_RESV,
+	__RQF_BITS
+};
+
+#define RQF_STARTED		((__force req_flags_t)(1 << __RQF_STARTED))
+#define RQF_FLUSH_SEQ		((__force req_flags_t)(1 << __RQF_FLUSH_SEQ))
+#define RQF_MIXED_MERGE		((__force req_flags_t)(1 << __RQF_MIXED_MERGE))
+#define RQF_DONTPREP		((__force req_flags_t)(1 << __RQF_DONTPREP))
+#define RQF_SCHED_TAGS		((__force req_flags_t)(1 << __RQF_SCHED_TAGS))
+#define RQF_USE_SCHED		((__force req_flags_t)(1 << __RQF_USE_SCHED))
+#define RQF_FAILED		((__force req_flags_t)(1 << __RQF_FAILED))
+#define RQF_QUIET		((__force req_flags_t)(1 << __RQF_QUIET))
+#define RQF_IO_STAT		((__force req_flags_t)(1 << __RQF_IO_STAT))
+#define RQF_PM			((__force req_flags_t)(1 << __RQF_PM))
+#define RQF_HASHED		((__force req_flags_t)(1 << __RQF_HASHED))
+#define RQF_STATS		((__force req_flags_t)(1 << __RQF_STATS))
+#define RQF_SPECIAL_PAYLOAD	\
+			((__force req_flags_t)(1 << __RQF_SPECIAL_PAYLOAD))
+#define RQF_ZONE_WRITE_PLUGGING	\
+			((__force req_flags_t)(1 << __RQF_ZONE_WRITE_PLUGGING))
+#define RQF_TIMED_OUT		((__force req_flags_t)(1 << __RQF_TIMED_OUT))
+#define RQF_RESV		((__force req_flags_t)(1 << __RQF_RESV))
 
 /* flags that prevent us from merging requests: */
 #define RQF_NOMERGE_FLAGS \
@@ -278,8 +301,12 @@ enum blk_eh_timer_return {
 	BLK_EH_RESET_TIMER,
 };
 
-#define BLK_TAG_ALLOC_FIFO 0 /* allocate starting from 0 */
-#define BLK_TAG_ALLOC_RR 1 /* allocate starting from last allocated tag */
+/* Keep alloc_policy_name[] in sync with the definitions below */
+enum {
+	BLK_TAG_ALLOC_FIFO,	/* allocate starting from 0 */
+	BLK_TAG_ALLOC_RR,	/* allocate starting from last allocated tag */
+	BLK_TAG_ALLOC_MAX
+};
 
 /**
  * struct blk_mq_hw_ctx - State for a hardware queue facing the hardware
@@ -644,6 +671,7 @@ struct blk_mq_ops {
 #endif
 };
 
+/* Keep hctx_flag_name[] in sync with the definitions below */
 enum {
 	BLK_MQ_F_SHOULD_MERGE	= 1 << 0,
 	BLK_MQ_F_TAG_QUEUE_SHARED = 1 << 1,
@@ -653,27 +681,17 @@ enum {
 	 */
 	BLK_MQ_F_STACKING	= 1 << 2,
 	BLK_MQ_F_TAG_HCTX_SHARED = 1 << 3,
-	BLK_MQ_F_BLOCKING	= 1 << 5,
+	BLK_MQ_F_BLOCKING	= 1 << 4,
 	/* Do not allow an I/O scheduler to be configured. */
-	BLK_MQ_F_NO_SCHED	= 1 << 6,
+	BLK_MQ_F_NO_SCHED	= 1 << 5,
+
 	/*
 	 * Select 'none' during queue registration in case of a single hwq
 	 * or shared hwqs instead of 'mq-deadline'.
 	 */
-	BLK_MQ_F_NO_SCHED_BY_DEFAULT	= 1 << 7,
-	BLK_MQ_F_ALLOC_POLICY_START_BIT = 8,
+	BLK_MQ_F_NO_SCHED_BY_DEFAULT	= 1 << 6,
+	BLK_MQ_F_ALLOC_POLICY_START_BIT = 7,
 	BLK_MQ_F_ALLOC_POLICY_BITS = 1,
-
-	BLK_MQ_S_STOPPED	= 0,
-	BLK_MQ_S_TAG_ACTIVE	= 1,
-	BLK_MQ_S_SCHED_RESTART	= 2,
-
-	/* hw queue is inactive after all its CPUs become offline */
-	BLK_MQ_S_INACTIVE	= 3,
-
-	BLK_MQ_MAX_DEPTH	= 10240,
-
-	BLK_MQ_CPU_WORK_BATCH	= 8,
 };
 #define BLK_MQ_FLAG_TO_ALLOC_POLICY(flags) \
 	((flags >> BLK_MQ_F_ALLOC_POLICY_START_BIT) & \
@@ -682,8 +700,19 @@ enum {
 	((policy & ((1 << BLK_MQ_F_ALLOC_POLICY_BITS) - 1)) \
 		<< BLK_MQ_F_ALLOC_POLICY_START_BIT)
 
+#define BLK_MQ_MAX_DEPTH	(10240)
 #define BLK_MQ_NO_HCTX_IDX	(-1U)
 
+enum {
+	/* Keep hctx_state_name[] in sync with the definitions below */
+	BLK_MQ_S_STOPPED,
+	BLK_MQ_S_TAG_ACTIVE,
+	BLK_MQ_S_SCHED_RESTART,
+	/* hw queue is inactive after all its CPUs become offline */
+	BLK_MQ_S_INACTIVE,
+	BLK_MQ_S_MAX
+};
+
 struct gendisk *__blk_mq_alloc_disk(struct blk_mq_tag_set *set,
 		struct queue_limits *lim, void *queuedata,
 		struct lock_class_key *lkclass);

include/linux/blk_types.h

@@ -354,6 +354,7 @@ enum req_op {
 	REQ_OP_LAST		= (__force blk_opf_t)36,
 };
 
+/* Keep cmd_flag_name[] in sync with the definitions below */
 enum req_flag_bits {
 	__REQ_FAILFAST_DEV =	/* no driver retries of device errors */
 		REQ_OP_BITS,

include/linux/blkdev.h

@@ -588,27 +588,28 @@ struct request_queue {
 };
 
 /* Keep blk_queue_flag_name[] in sync with the definitions below */
-#define QUEUE_FLAG_STOPPED	0	/* queue is stopped */
-#define QUEUE_FLAG_DYING	1	/* queue being torn down */
-#define QUEUE_FLAG_NOMERGES     3	/* disable merge attempts */
-#define QUEUE_FLAG_SAME_COMP	4	/* complete on same CPU-group */
-#define QUEUE_FLAG_FAIL_IO	5	/* fake timeout */
-#define QUEUE_FLAG_NOXMERGES	9	/* No extended merges */
-#define QUEUE_FLAG_SAME_FORCE	12	/* force complete on same CPU */
-#define QUEUE_FLAG_INIT_DONE	14	/* queue is initialized */
-#define QUEUE_FLAG_STATS	20	/* track IO start and completion times */
-#define QUEUE_FLAG_REGISTERED	22	/* queue has been registered to a disk */
-#define QUEUE_FLAG_QUIESCED	24	/* queue has been quiesced */
-#define QUEUE_FLAG_RQ_ALLOC_TIME 27	/* record rq->alloc_time_ns */
-#define QUEUE_FLAG_HCTX_ACTIVE	28	/* at least one blk-mq hctx is active */
-#define QUEUE_FLAG_SQ_SCHED     30	/* single queue style io dispatch */
+enum {
+	QUEUE_FLAG_DYING,		/* queue being torn down */
+	QUEUE_FLAG_NOMERGES,		/* disable merge attempts */
+	QUEUE_FLAG_SAME_COMP,		/* complete on same CPU-group */
+	QUEUE_FLAG_FAIL_IO,		/* fake timeout */
+	QUEUE_FLAG_NOXMERGES,		/* No extended merges */
+	QUEUE_FLAG_SAME_FORCE,		/* force complete on same CPU */
+	QUEUE_FLAG_INIT_DONE,		/* queue is initialized */
+	QUEUE_FLAG_STATS,		/* track IO start and completion times */
+	QUEUE_FLAG_REGISTERED,		/* queue has been registered to a disk */
+	QUEUE_FLAG_QUIESCED,		/* queue has been quiesced */
+	QUEUE_FLAG_RQ_ALLOC_TIME,	/* record rq->alloc_time_ns */
+	QUEUE_FLAG_HCTX_ACTIVE,		/* at least one blk-mq hctx is active */
+	QUEUE_FLAG_SQ_SCHED,		/* single queue style io dispatch */
+	QUEUE_FLAG_MAX
+};
 
 #define QUEUE_FLAG_MQ_DEFAULT	(1UL << QUEUE_FLAG_SAME_COMP)
 
 void blk_queue_flag_set(unsigned int flag, struct request_queue *q);
 void blk_queue_flag_clear(unsigned int flag, struct request_queue *q);
 
-#define blk_queue_stopped(q)	test_bit(QUEUE_FLAG_STOPPED, &(q)->queue_flags)
 #define blk_queue_dying(q)	test_bit(QUEUE_FLAG_DYING, &(q)->queue_flags)
 #define blk_queue_init_done(q)	test_bit(QUEUE_FLAG_INIT_DONE, &(q)->queue_flags)
 #define blk_queue_nomerges(q)	test_bit(QUEUE_FLAG_NOMERGES, &(q)->queue_flags)

include/linux/cgroup-defs.h

@@ -539,9 +539,6 @@ struct cgroup {
 	/* used to store eBPF programs */
 	struct cgroup_bpf bpf;
 
-	/* If there is block congestion on this cgroup. */
-	atomic_t congestion_count;
-
 	/* Used to store internal freezer state */
 	struct cgroup_freezer_state freezer;
 

include/linux/sbitmap.h

@@ -36,6 +36,11 @@ struct sbitmap_word {
 	 * @cleared: word holding cleared bits
 	 */
 	unsigned long cleared ____cacheline_aligned_in_smp;
+
+	/**
+	 * @swap_lock: serializes simultaneous updates of ->word and ->cleared
+	 */
+	spinlock_t swap_lock;
 } ____cacheline_aligned_in_smp;
 
 /**

lib/sbitmap.c

@@ -60,12 +60,30 @@ static inline void update_alloc_hint_after_get(struct sbitmap *sb,
 /*
  * See if we have deferred clears that we can batch move
  */
-static inline bool sbitmap_deferred_clear(struct sbitmap_word *map)
+static inline bool sbitmap_deferred_clear(struct sbitmap_word *map,
+		unsigned int depth, unsigned int alloc_hint, bool wrap)
 {
-	unsigned long mask;
+	unsigned long mask, word_mask;
 
-	if (!READ_ONCE(map->cleared))
-		return false;
+	guard(spinlock_irqsave)(&map->swap_lock);
+
+	if (!map->cleared) {
+		if (depth == 0)
+			return false;
+
+		word_mask = (~0UL) >> (BITS_PER_LONG - depth);
+		/*
+		 * The current behavior is to always retry after moving
+		 * ->cleared to word, and we change it to retry in case
+		 * of any free bits. To avoid an infinite loop, we need
+		 * to take wrap & alloc_hint into account, otherwise a
+		 * soft lockup may occur.
+		 */
+		if (!wrap && alloc_hint)
+			word_mask &= ~((1UL << alloc_hint) - 1);
+
+		return (READ_ONCE(map->word) & word_mask) != word_mask;
+	}
 
 	/*
 	 * First get a stable cleared mask, setting the old mask to 0.
@@ -85,6 +103,7 @@ int sbitmap_init_node(struct sbitmap *sb, unsigned int depth, int shift,
 		      bool alloc_hint)
 {
 	unsigned int bits_per_word;
+	int i;
 
 	if (shift < 0)
 		shift = sbitmap_calculate_shift(depth);
@@ -116,6 +135,9 @@ int sbitmap_init_node(struct sbitmap *sb, unsigned int depth, int shift,
 		return -ENOMEM;
 	}
 
+	for (i = 0; i < sb->map_nr; i++)
+		spin_lock_init(&sb->map[i].swap_lock);
+
 	return 0;
 }
 EXPORT_SYMBOL_GPL(sbitmap_init_node);
@@ -126,7 +148,7 @@ void sbitmap_resize(struct sbitmap *sb, unsigned int depth)
 	unsigned int i;
 
 	for (i = 0; i < sb->map_nr; i++)
-		sbitmap_deferred_clear(&sb->map[i]);
+		sbitmap_deferred_clear(&sb->map[i], 0, 0, 0);
 
 	sb->depth = depth;
 	sb->map_nr = DIV_ROUND_UP(sb->depth, bits_per_word);
@@ -179,7 +201,7 @@ static int sbitmap_find_bit_in_word(struct sbitmap_word *map,
 					alloc_hint, wrap);
 		if (nr != -1)
 			break;
-		if (!sbitmap_deferred_clear(map))
+		if (!sbitmap_deferred_clear(map, depth, alloc_hint, wrap))
 			break;
 	} while (1);
 
@@ -496,7 +518,7 @@ unsigned long __sbitmap_queue_get_batch(struct sbitmap_queue *sbq, int nr_tags,
 		unsigned int map_depth = __map_depth(sb, index);
 		unsigned long val;
 
-		sbitmap_deferred_clear(map);
+		sbitmap_deferred_clear(map, 0, 0, 0);
 		val = READ_ONCE(map->word);
 		if (val == (1UL << (map_depth - 1)) - 1)
 			goto next;