public inbox for linux-block@vger.kernel.org
 help / color / mirror / Atom feed
From: Ming Lei <ming.lei@redhat.com>
To: Jens Axboe <axboe@kernel.dk>, linux-block@vger.kernel.org
Cc: "Nilay Shroff" <nilay@linux.ibm.com>,
	"Shinichiro Kawasaki" <shinichiro.kawasaki@wdc.com>,
	"Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	"Christoph Hellwig" <hch@lst.de>,
	"Ming Lei" <ming.lei@redhat.com>
Subject: [PATCH V3 07/20] block: prevent adding/deleting disk during updating nr_hw_queues
Date: Thu, 24 Apr 2025 23:21:30 +0800	[thread overview]
Message-ID: <20250424152148.1066220-8-ming.lei@redhat.com> (raw)
In-Reply-To: <20250424152148.1066220-1-ming.lei@redhat.com>

Both adding/deleting disk code are reader of `nr_hw_queues`, so we can't
allow them in-progress when updating nr_hw_queues, kernel panic and
kasan has been reported in [1].

Prevent adding/deleting disk during updating nr_hw_queues by adding
rw_semaphore to tagset, write lock is grabbed in blk_mq_update_nr_hw_queues(),
and read lock is acquired when adding/deleting disk.

Also mark GFP_NOIO allocation scope for adding/deleting disk because
blk_mq_update_nr_hw_queues() is part of some driver's error handler.

This way avoids lot of trouble.

Suggested-by: Nilay Shroff <nilay@linux.ibm.com>
Reported-by: Nilay Shroff <nilay@linux.ibm.com>
Closes: https://lore.kernel.org/linux-block/a5896cdb-a59a-4a37-9f99-20522f5d2987@linux.ibm.com/
Signed-off-by: Ming Lei <ming.lei@redhat.com>
---
 block/blk-mq.c         |  4 ++
 block/genhd.c          | 94 +++++++++++++++++++++++++++++++-----------
 include/linux/blk-mq.h |  3 ++
 3 files changed, 78 insertions(+), 23 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 29cfc7ce2e0a..1ed2d183f912 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -4802,6 +4802,8 @@ int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set)
 			goto out_free_srcu;
 	}
 
+	init_rwsem(&set->update_nr_hwq_sema);
+
 	ret = -ENOMEM;
 	set->tags = kcalloc_node(set->nr_hw_queues,
 				 sizeof(struct blk_mq_tags *), GFP_KERNEL,
@@ -5097,9 +5099,11 @@ static void __blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set,
 
 void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues)
 {
+	down_write(&set->update_nr_hwq_sema);
 	mutex_lock(&set->tag_list_lock);
 	__blk_mq_update_nr_hw_queues(set, nr_hw_queues);
 	mutex_unlock(&set->tag_list_lock);
+	up_write(&set->update_nr_hwq_sema);
 }
 EXPORT_SYMBOL_GPL(blk_mq_update_nr_hw_queues);
 
diff --git a/block/genhd.c b/block/genhd.c
index c2bd86cd09de..7f3ae3d23b26 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -399,9 +399,9 @@ int disk_scan_partitions(struct gendisk *disk, blk_mode_t mode)
  * This function registers the partitioning information in @disk
  * with the kernel. Also attach a fwnode to the disk device.
  */
-int __must_check add_disk_fwnode(struct device *parent, struct gendisk *disk,
-				 const struct attribute_group **groups,
-				 struct fwnode_handle *fwnode)
+static int __add_disk_fwnode(struct device *parent, struct gendisk *disk,
+			     const struct attribute_group **groups,
+			     struct fwnode_handle *fwnode)
 
 {
 	struct device *ddev = disk_to_dev(disk);
@@ -572,6 +572,37 @@ int __must_check add_disk_fwnode(struct device *parent, struct gendisk *disk,
 	}
 	return ret;
 }
+
+/**
+ * add_disk_fwnode - add disk information to kernel list with fwnode
+ * @parent: parent device for the disk
+ * @disk: per-device partitioning information
+ * @groups: Additional per-device sysfs groups
+ * @fwnode: attached disk fwnode
+ *
+ * This function registers the partitioning information in @disk
+ * with the kernel. Also attach a fwnode to the disk device.
+ */
+int __must_check add_disk_fwnode(struct device *parent, struct gendisk *disk,
+				 const struct attribute_group **groups,
+				 struct fwnode_handle *fwnode)
+{
+	struct blk_mq_tag_set *set;
+	unsigned int memflags;
+	int ret;
+
+	if (!queue_is_mq(disk->queue))
+		return __add_disk_fwnode(parent, disk, groups, fwnode);
+
+	set = disk->queue->tag_set;
+	memflags = memalloc_noio_save();
+	down_read(&set->update_nr_hwq_sema);
+	ret = __add_disk_fwnode(parent, disk, groups, fwnode);
+	up_read(&set->update_nr_hwq_sema);
+	memalloc_noio_restore(memflags);
+
+	return ret;
+}
 EXPORT_SYMBOL_GPL(add_disk_fwnode);
 
 /**
@@ -652,26 +683,7 @@ void blk_mark_disk_dead(struct gendisk *disk)
 }
 EXPORT_SYMBOL_GPL(blk_mark_disk_dead);
 
-/**
- * del_gendisk - remove the gendisk
- * @disk: the struct gendisk to remove
- *
- * Removes the gendisk and all its associated resources. This deletes the
- * partitions associated with the gendisk, and unregisters the associated
- * request_queue.
- *
- * This is the counter to the respective __device_add_disk() call.
- *
- * The final removal of the struct gendisk happens when its refcount reaches 0
- * with put_disk(), which should be called after del_gendisk(), if
- * __device_add_disk() was used.
- *
- * Drivers exist which depend on the release of the gendisk to be synchronous,
- * it should not be deferred.
- *
- * Context: can sleep
- */
-void del_gendisk(struct gendisk *disk)
+static void __del_gendisk(struct gendisk *disk)
 {
 	struct request_queue *q = disk->queue;
 	struct block_device *part;
@@ -764,6 +776,42 @@ void del_gendisk(struct gendisk *disk)
 	if (start_drain)
 		blk_unfreeze_release_lock(q);
 }
+
+/**
+ * del_gendisk - remove the gendisk
+ * @disk: the struct gendisk to remove
+ *
+ * Removes the gendisk and all its associated resources. This deletes the
+ * partitions associated with the gendisk, and unregisters the associated
+ * request_queue.
+ *
+ * This is the counter to the respective __device_add_disk() call.
+ *
+ * The final removal of the struct gendisk happens when its refcount reaches 0
+ * with put_disk(), which should be called after del_gendisk(), if
+ * __device_add_disk() was used.
+ *
+ * Drivers exist which depend on the release of the gendisk to be synchronous,
+ * it should not be deferred.
+ *
+ * Context: can sleep
+ */
+void del_gendisk(struct gendisk *disk)
+{
+	struct blk_mq_tag_set *set;
+	unsigned int memflags;
+
+	if (!queue_is_mq(disk->queue)) {
+		__del_gendisk(disk);
+	} else {
+		set = disk->queue->tag_set;
+		memflags = memalloc_noio_save();
+		down_read(&set->update_nr_hwq_sema);
+		__del_gendisk(disk);
+		up_read(&set->update_nr_hwq_sema);
+		memalloc_noio_restore(memflags);
+	}
+}
 EXPORT_SYMBOL(del_gendisk);
 
 /**
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index 8eb9b3310167..28bc03b2b0dc 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -9,6 +9,7 @@
 #include <linux/prefetch.h>
 #include <linux/srcu.h>
 #include <linux/rw_hint.h>
+#include <linux/rwsem.h>
 
 struct blk_mq_tags;
 struct blk_flush_queue;
@@ -527,6 +528,8 @@ struct blk_mq_tag_set {
 	struct mutex		tag_list_lock;
 	struct list_head	tag_list;
 	struct srcu_struct	*srcu;
+
+	struct rw_semaphore	update_nr_hwq_sema;
 };
 
 /**
-- 
2.47.0


  parent reply	other threads:[~2025-04-24 15:22 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-24 15:21 [PATCH V3 00/20] block: unify elevator changing and fix lockdep warning Ming Lei
2025-04-24 15:21 ` [PATCH V3 01/20] block: move blk_mq_add_queue_tag_set() after blk_mq_map_swqueue() Ming Lei
2025-04-24 15:21 ` [PATCH V3 02/20] block: move ELEVATOR_FLAG_DISABLE_WBT a request queue flag Ming Lei
2025-04-25 14:29   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 03/20] block: don't call freeze queue in elevator_switch() and elevator_disable() Ming Lei
2025-04-25 14:29   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 04/20] block: use q->elevator with ->elevator_lock held in elv_iosched_show() Ming Lei
2025-04-25  6:08   ` Hannes Reinecke
2025-04-25 10:53   ` Nilay Shroff
2025-04-25 14:29   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 05/20] block: add two helpers for registering/un-registering sched debugfs Ming Lei
2025-04-24 15:21 ` [PATCH V3 06/20] block: move sched debugfs register into elvevator_register_queue Ming Lei
2025-04-24 15:21 ` Ming Lei [this message]
2025-04-25  6:33   ` [PATCH V3 07/20] block: prevent adding/deleting disk during updating nr_hw_queues Hannes Reinecke
2025-04-25 11:37   ` Nilay Shroff
2025-04-25 14:33   ` Christoph Hellwig
2025-04-25 16:46   ` kernel test robot
2025-04-24 15:21 ` [PATCH V3 08/20] block: don't allow to switch elevator if updating nr_hw_queues is in-progress Ming Lei
2025-04-25  6:33   ` Hannes Reinecke
2025-04-25 12:48   ` Nilay Shroff
2025-04-27  2:27     ` Ming Lei
2025-04-28 16:17       ` Nilay Shroff
2025-04-29  2:43         ` Ming Lei
2025-04-29 10:22           ` Nilay Shroff
2025-04-30  0:54             ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 09/20] block: simplify elevator reattachment for updating nr_hw_queues Ming Lei
2025-04-25  6:34   ` Hannes Reinecke
2025-04-25 18:12   ` Christoph Hellwig
2025-04-29  9:51     ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 10/20] block: move blk_unregister_queue() & device_del() after freeze wait Ming Lei
2025-04-25  6:35   ` Hannes Reinecke
2025-04-25 12:50   ` Nilay Shroff
2025-04-25 14:34   ` Christoph Hellwig
2025-04-28 11:51     ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 11/20] block: move queue freezing & elevator_lock into elevator_change() Ming Lei
2025-04-25  6:36   ` Hannes Reinecke
2025-04-25 12:54   ` Nilay Shroff
2025-04-25 14:35   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 12/20] block: add `struct elv_change_ctx` for unifying elevator change Ming Lei
2025-04-25  6:38   ` Hannes Reinecke
2025-04-25 18:23   ` Christoph Hellwig
2025-04-29 15:45     ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 13/20] block: " Ming Lei
2025-04-25  6:39   ` Hannes Reinecke
2025-04-25 13:03   ` Nilay Shroff
2025-04-30  0:46     ` Ming Lei
2025-04-25 18:30   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 14/20] block: pass elevator_queue to elv_register_queue & unregister_queue Ming Lei
2025-04-25  6:40   ` Hannes Reinecke
2025-04-24 15:21 ` [PATCH V3 15/20] block: fail to show/store elevator sysfs attribute if elevator is dying Ming Lei
2025-04-25  6:45   ` Hannes Reinecke
2025-04-25 18:36   ` Christoph Hellwig
2025-04-30  1:07     ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 16/20] block: move elv_register[unregister]_queue out of elevator_lock Ming Lei
2025-04-25  6:46   ` Hannes Reinecke
2025-04-25 13:05   ` Nilay Shroff
2025-04-25 18:37   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 17/20] block: move debugfs/sysfs register out of freezing queue Ming Lei
2025-04-25  6:47   ` Hannes Reinecke
2025-04-25 18:38   ` Christoph Hellwig
2025-04-28 11:28     ` Ming Lei
2025-04-24 15:21 ` [PATCH V3 18/20] block: remove several ->elevator_lock Ming Lei
2025-04-25  6:48   ` Hannes Reinecke
2025-04-25 18:38   ` Christoph Hellwig
2025-04-24 15:21 ` [PATCH V3 19/20] block: move hctx cpuhp add/del out of queue freezing Ming Lei
2025-04-25  6:49   ` Hannes Reinecke
2025-04-24 15:21 ` [PATCH V3 20/20] block: move wbt_enable_default() out of queue freezing from sched ->exit() Ming Lei
2025-04-25 13:10   ` Nilay Shroff
2025-04-29 10:59   ` Nilay Shroff
2025-04-29 12:00 ` [PATCH V3 00/20] block: unify elevator changing and fix lockdep warning Stefan Haberland
2025-04-29 12:11   ` Ming Lei

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250424152148.1066220-8-ming.lei@redhat.com \
    --to=ming.lei@redhat.com \
    --cc=axboe@kernel.dk \
    --cc=hch@lst.de \
    --cc=linux-block@vger.kernel.org \
    --cc=nilay@linux.ibm.com \
    --cc=shinichiro.kawasaki@wdc.com \
    --cc=thomas.hellstrom@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox