Commit d1cf752d authored by Linus Torvalds's avatar Linus Torvalds
Browse files

Merge tag 'block-6.17-20250828' of git://git.kernel.dk/linux 

Pull block fixes from Jens Axboe:

 - Fix a lockdep spotted issue on recursive locking for zoned writes, in
   case of errors

 - Update bcache MAINTAINERS entry address for Coly

 - Fix for a ublk release issue, with selftests

 - Fix for a regression introduced in this cycle, where it assumed
   q->rq_qos was always set if the bio flag indicated that

 - Fix for a regression introduced in this cycle, where loop retrieving
   block device sizes got broken

* tag 'block-6.17-20250828' of git://git.kernel.dk/linux:
  bcache: change maintainer's email address
  ublk selftests: add --no_ublk_fixed_fd for not using registered ublk char device
  ublk: avoid ublk_io_release() called after ublk char dev is closed
  block: validate QoS before calling __rq_qos_done_bio()
  blk-zoned: Fix a lockdep complaint about recursive locking
  loop: fix zero sized loop for block special file
parents 220374ab 95a7c500

MAINTAINERS

+1 −1
Original line number Diff line number Diff line
@@ -4205,7 +4205,7 @@ W: http://www.baycom.org/~tom/ham/ham.html 
F:	drivers/net/hamradio/baycom*









BCACHE (BLOCK LAYER CACHE)









M:	Coly Li <colyli@kernel.org>









M:	Coly Li <colyli@fnnas.com>









M:	Kent Overstreet <kent.overstreet@linux.dev>









L:	linux-bcache@vger.kernel.org









S:	Maintained
































block/blk-rq-qos.h



+8
−5






























Original line number
Diff line number
Diff line







@@ -149,11 +149,14 @@ static inline void rq_qos_done_bio(struct bio *bio)







	q = bdev_get_queue(bio->bi_bdev);



















	/*









	 * If a bio has BIO_QOS_xxx set, it implicitly implies that









	 * q->rq_qos is present. So, we skip re-checking q->rq_qos









	 * here as an extra optimization and directly call









	 * __rq_qos_done_bio().









	 * A BIO may carry BIO_QOS_* flags even if the associated request_queue









	 * does not have rq_qos enabled. This can happen with stacked block









	 * devices — for example, NVMe multipath, where it's possible that the









	 * bottom device has QoS enabled but the top device does not. Therefore,









	 * always verify that q->rq_qos is present and QoS is enabled before









	 * calling __rq_qos_done_bio().









	 */









	if (test_bit(QUEUE_FLAG_QOS_ENABLED, &q->queue_flags) && q->rq_qos)









		__rq_qos_done_bio(q->rq_qos, bio);









}










































block/blk-zoned.c



+6
−5






























Original line number
Diff line number
Diff line







@@ -1286,14 +1286,14 @@ static void blk_zone_wplug_bio_work(struct work_struct *work)







	struct block_device *bdev;









	unsigned long flags;









	struct bio *bio;









	bool prepared;



















	/*









	 * Submit the next plugged BIO. If we do not have any, clear









	 * the plugged flag.









	 */









	spin_lock_irqsave(&zwplug->lock, flags);



















again:









	spin_lock_irqsave(&zwplug->lock, flags);









	bio = bio_list_pop(&zwplug->bio_list);









	if (!bio) {









		zwplug->flags &= ~BLK_ZONE_WPLUG_PLUGGED;








@@ -1304,13 +1304,14 @@ static void blk_zone_wplug_bio_work(struct work_struct *work)







	trace_blk_zone_wplug_bio(zwplug->disk->queue, zwplug->zone_no,









				 bio->bi_iter.bi_sector, bio_sectors(bio));



















	if (!blk_zone_wplug_prepare_bio(zwplug, bio)) {









	prepared = blk_zone_wplug_prepare_bio(zwplug, bio);









	spin_unlock_irqrestore(&zwplug->lock, flags);



















	if (!prepared) {









		blk_zone_wplug_bio_io_error(zwplug, bio);









		goto again;









	}



















	spin_unlock_irqrestore(&zwplug->lock, flags);



















	bdev = bio->bi_bdev;



















	/*
































drivers/block/loop.c



+16
−10






























Original line number
Diff line number
Diff line







@@ -139,10 +139,14 @@ static int part_shift;

















static loff_t lo_calculate_size(struct loop_device *lo, struct file *file)









{









	struct kstat stat;









	loff_t loopsize;









	int ret;



















	if (S_ISBLK(file_inode(file)->i_mode)) {









		loopsize = i_size_read(file->f_mapping->host);









	} else {









		struct kstat stat;



















		/*









		 * Get the accurate file size. This provides better results than









		 * cached inode data, particularly for network filesystems where








@@ -153,6 +157,8 @@ static loff_t lo_calculate_size(struct loop_device *lo, struct file *file)







			return 0;



















		loopsize = stat.size;









	}



















	if (lo->lo_offset > 0)









		loopsize -= lo->lo_offset;









	/* offset is beyond i_size, weird but possible */
































drivers/block/ublk_drv.c



+70
−2






























Original line number
Diff line number
Diff line







@@ -239,6 +239,7 @@ struct ublk_device {







	struct mutex cancel_mutex;









	bool canceling;









	pid_t 	ublksrv_tgid;









	struct delayed_work	exit_work;









};



















/* header of ublk_params */








@@ -1595,12 +1596,62 @@ static void ublk_set_canceling(struct ublk_device *ub, bool canceling)







		ublk_get_queue(ub, i)->canceling = canceling;









}



















static int ublk_ch_release(struct inode *inode, struct file *filp)









static bool ublk_check_and_reset_active_ref(struct ublk_device *ub)









{









	struct ublk_device *ub = filp->private_data;









	int i, j;



















	if (!(ub->dev_info.flags & (UBLK_F_SUPPORT_ZERO_COPY |









					UBLK_F_AUTO_BUF_REG)))









		return false;



















	for (i = 0; i < ub->dev_info.nr_hw_queues; i++) {









		struct ublk_queue *ubq = ublk_get_queue(ub, i);



















		for (j = 0; j < ubq->q_depth; j++) {









			struct ublk_io *io = &ubq->ios[j];









			unsigned int refs = refcount_read(&io->ref) +









				io->task_registered_buffers;



















			/*









			 * UBLK_REFCOUNT_INIT or zero means no active









			 * reference









			 */









			if (refs != UBLK_REFCOUNT_INIT && refs != 0)









				return true;



















			/* reset to zero if the io hasn't active references */









			refcount_set(&io->ref, 0);









			io->task_registered_buffers = 0;









		}









	}









	return false;









}



















static void ublk_ch_release_work_fn(struct work_struct *work)









{









	struct ublk_device *ub =









		container_of(work, struct ublk_device, exit_work.work);









	struct gendisk *disk;









	int i;



















	/*









	 * For zero-copy and auto buffer register modes, I/O references









	 * might not be dropped naturally when the daemon is killed, but









	 * io_uring guarantees that registered bvec kernel buffers are









	 * unregistered finally when freeing io_uring context, then the









	 * active references are dropped.









	 *









	 * Wait until active references are dropped for avoiding use-after-free









	 *









	 * registered buffer may be unregistered in io_ring's release hander,









	 * so have to wait by scheduling work function for avoiding the two









	 * file release dependency.









	 */









	if (ublk_check_and_reset_active_ref(ub)) {









		schedule_delayed_work(&ub->exit_work, 1);









		return;









	}



















	/*









	 * disk isn't attached yet, either device isn't live, or it has









	 * been removed already, so we needn't to do anything








@@ -1673,6 +1724,23 @@ static int ublk_ch_release(struct inode *inode, struct file *filp)







	ublk_reset_ch_dev(ub);









out:









	clear_bit(UB_STATE_OPEN, &ub->state);



















	/* put the reference grabbed in ublk_ch_release() */









	ublk_put_device(ub);









}



















static int ublk_ch_release(struct inode *inode, struct file *filp)









{









	struct ublk_device *ub = filp->private_data;



















	/*









	 * Grab ublk device reference, so it won't be gone until we are









	 * really released from work function.









	 */









	ublk_get_device(ub);



















	INIT_DELAYED_WORK(&ub->exit_work, ublk_ch_release_work_fn);









	schedule_delayed_work(&ub->exit_work, 0);









	return 0;









}