Merge tag 'block-6.13-20241220' of git://git.kernel.dk/linux

Pull block fixes from Jens Axboe:

 - Minor cleanups for bdev/nvme using the helpers introduced

 - Revert of a deadlock fix that still needs more work

 - Fix a UAF of hctx in the cpu hotplug code

* tag 'block-6.13-20241220' of git://git.kernel.dk/linux:
  block: avoid to reuse `hctx` not removed from cpuhp callback list
  block: Revert "block: Fix potential deadlock while freezing queue and acquiring sysfs_lock"
  nvme: use blk_validate_block_size() for max LBA check
  block/bdev: use helper for max block size check

Author: torvalds

block/bdev.c

@@ -155,8 +155,7 @@ int set_blocksize(struct file *file, int size)
 	struct inode *inode = file->f_mapping->host;
 	struct block_device *bdev = I_BDEV(inode);
 
-	/* Size must be a power of two, and between 512 and PAGE_SIZE */
-	if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size))
+	if (blk_validate_block_size(size))
 		return -EINVAL;
 
 	/* Size cannot be smaller than the size supported by the device */

block/blk-mq-sysfs.c

@@ -275,13 +275,15 @@ void blk_mq_sysfs_unregister_hctxs(struct request_queue *q)
 	struct blk_mq_hw_ctx *hctx;
 	unsigned long i;
 
-	lockdep_assert_held(&q->sysfs_dir_lock);
-
+	mutex_lock(&q->sysfs_dir_lock);
 	if (!q->mq_sysfs_init_done)
-		return;
+		goto unlock;
 
 	queue_for_each_hw_ctx(q, hctx, i)
 		blk_mq_unregister_hctx(hctx);
+
+unlock:
+	mutex_unlock(&q->sysfs_dir_lock);
 }
 
 int blk_mq_sysfs_register_hctxs(struct request_queue *q)
@@ -290,16 +292,18 @@ int blk_mq_sysfs_register_hctxs(struct request_queue *q)
 	unsigned long i;
 	int ret = 0;
 
-	lockdep_assert_held(&q->sysfs_dir_lock);
-
+	mutex_lock(&q->sysfs_dir_lock);
 	if (!q->mq_sysfs_init_done)
-		return ret;
+		goto unlock;
 
 	queue_for_each_hw_ctx(q, hctx, i) {
 		ret = blk_mq_register_hctx(hctx);
 		if (ret)
 			break;
 	}
 
+unlock:
+	mutex_unlock(&q->sysfs_dir_lock);
+
 	return ret;
 }

block/blk-mq.c

@@ -4412,6 +4412,15 @@ struct gendisk *blk_mq_alloc_disk_for_queue(struct request_queue *q,
 }
 EXPORT_SYMBOL(blk_mq_alloc_disk_for_queue);
 
+/*
+ * Only hctx removed from cpuhp list can be reused
+ */
+static bool blk_mq_hctx_is_reusable(struct blk_mq_hw_ctx *hctx)
+{
+	return hlist_unhashed(&hctx->cpuhp_online) &&
+		hlist_unhashed(&hctx->cpuhp_dead);
+}
+
 static struct blk_mq_hw_ctx *blk_mq_alloc_and_init_hctx(
 		struct blk_mq_tag_set *set, struct request_queue *q,
 		int hctx_idx, int node)
@@ -4421,7 +4430,7 @@ static struct blk_mq_hw_ctx *blk_mq_alloc_and_init_hctx(
 	/* reuse dead hctx first */
 	spin_lock(&q->unused_hctx_lock);
 	list_for_each_entry(tmp, &q->unused_hctx_list, hctx_list) {
-		if (tmp->numa_node == node) {
+		if (tmp->numa_node == node && blk_mq_hctx_is_reusable(tmp)) {
 			hctx = tmp;
 			break;
 		}
@@ -4453,8 +4462,7 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set,
 	unsigned long i, j;
 
 	/* protect against switching io scheduler  */
-	lockdep_assert_held(&q->sysfs_lock);
-
+	mutex_lock(&q->sysfs_lock);
 	for (i = 0; i < set->nr_hw_queues; i++) {
 		int old_node;
 		int node = blk_mq_get_hctx_node(set, i);
@@ -4487,6 +4495,7 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set,
 
 	xa_for_each_start(&q->hctx_table, j, hctx, j)
 		blk_mq_exit_hctx(q, set, hctx, j);
+	mutex_unlock(&q->sysfs_lock);
 
 	/* unregister cpuhp callbacks for exited hctxs */
 	blk_mq_remove_hw_queues_cpuhp(q);
@@ -4518,14 +4527,10 @@ int blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
 
 	xa_init(&q->hctx_table);
 
-	mutex_lock(&q->sysfs_lock);
-
 	blk_mq_realloc_hw_ctxs(set, q);
 	if (!q->nr_hw_queues)
 		goto err_hctxs;
 
-	mutex_unlock(&q->sysfs_lock);
-
 	INIT_WORK(&q->timeout_work, blk_mq_timeout_work);
 	blk_queue_rq_timeout(q, set->timeout ? set->timeout : 30 * HZ);
 
@@ -4544,7 +4549,6 @@ int blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
 	return 0;
 
 err_hctxs:
-	mutex_unlock(&q->sysfs_lock);
 	blk_mq_release(q);
 err_exit:
 	q->mq_ops = NULL;
@@ -4925,12 +4929,12 @@ static bool blk_mq_elv_switch_none(struct list_head *head,
 		return false;
 
 	/* q->elevator needs protection from ->sysfs_lock */
-	lockdep_assert_held(&q->sysfs_lock);
+	mutex_lock(&q->sysfs_lock);
 
 	/* the check has to be done with holding sysfs_lock */
 	if (!q->elevator) {
 		kfree(qe);
-		goto out;
+		goto unlock;
 	}
 
 	INIT_LIST_HEAD(&qe->node);
@@ -4940,7 +4944,9 @@ static bool blk_mq_elv_switch_none(struct list_head *head,
 	__elevator_get(qe->type);
 	list_add(&qe->node, head);
 	elevator_disable(q);
-out:
+unlock:
+	mutex_unlock(&q->sysfs_lock);
+
 	return true;
 }
 
@@ -4969,9 +4975,11 @@ static void blk_mq_elv_switch_back(struct list_head *head,
 	list_del(&qe->node);
 	kfree(qe);
 
+	mutex_lock(&q->sysfs_lock);
 	elevator_switch(q, t);
 	/* drop the reference acquired in blk_mq_elv_switch_none */
 	elevator_put(t);
+	mutex_unlock(&q->sysfs_lock);
 }
 
 static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
@@ -4991,11 +4999,8 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
 	if (set->nr_maps == 1 && nr_hw_queues == set->nr_hw_queues)
 		return;
 
-	list_for_each_entry(q, &set->tag_list, tag_set_list) {
-		mutex_lock(&q->sysfs_dir_lock);
-		mutex_lock(&q->sysfs_lock);
+	list_for_each_entry(q, &set->tag_list, tag_set_list)
 		blk_mq_freeze_queue(q);
-	}
 	/*
 	 * Switch IO scheduler to 'none', cleaning up the data associated
 	 * with the previous scheduler. We will switch back once we are done
@@ -5051,11 +5056,8 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
 	list_for_each_entry(q, &set->tag_list, tag_set_list)
 		blk_mq_elv_switch_back(&head, q);
 
-	list_for_each_entry(q, &set->tag_list, tag_set_list) {
+	list_for_each_entry(q, &set->tag_list, tag_set_list)
 		blk_mq_unfreeze_queue(q);
-		mutex_unlock(&q->sysfs_lock);
-		mutex_unlock(&q->sysfs_dir_lock);
-	}
 
 	/* Free the excess tags when nr_hw_queues shrink. */
 	for (i = set->nr_hw_queues; i < prev_nr_hw_queues; i++)

block/blk-sysfs.c

@@ -706,11 +706,11 @@ queue_attr_store(struct kobject *kobj, struct attribute *attr,
 	if (entry->load_module)
 		entry->load_module(disk, page, length);
 
-	mutex_lock(&q->sysfs_lock);
 	blk_mq_freeze_queue(q);
+	mutex_lock(&q->sysfs_lock);
 	res = entry->store(disk, page, length);
-	blk_mq_unfreeze_queue(q);
 	mutex_unlock(&q->sysfs_lock);
+	blk_mq_unfreeze_queue(q);
 	return res;
 }
 

drivers/nvme/host/core.c

@@ -2034,7 +2034,7 @@ static bool nvme_update_disk_info(struct nvme_ns *ns, struct nvme_id_ns *id,
 	 * or smaller than a sector size yet, so catch this early and don't
 	 * allow block I/O.
 	 */
-	if (head->lba_shift > PAGE_SHIFT || head->lba_shift < SECTOR_SHIFT) {
+	if (blk_validate_block_size(bs)) {
 		bs = (1 << 9);
 		valid = false;
 	}