Merge tag 'block-5.14-2021-07-16' of git://git.kernel.dk/linux-block

Pull block fixes from Jens Axboe:

 - NVMe fixes via Christoph:
      - fix various races in nvme-pci when shutting down just after
        probing (Casey Chen)
      - fix a net_device leak in nvme-tcp (Prabhakar Kushwaha)

 - Fix regression in xen-blkfront by cleaning up the removal state
   machine (Christoph)

 - Fix tag_set and queue cleanup ordering regression in nbd (Wang)

 - Fix tag_set and queue cleanup ordering regression in pd (Guoqing)

* tag 'block-5.14-2021-07-16' of git://git.kernel.dk/linux-block:
  xen-blkfront: sanitize the removal state machine
  nbd: fix order of cleaning up the queue and freeing the tagset
  pd: fix order of cleaning up the queue and freeing the tagset
  nvme-pci: do not call nvme_dev_remove_admin from nvme_remove
  nvme-pci: fix multiple races in nvme_setup_io_queues
  nvme-tcp: use __dev_get_by_name instead dev_get_by_name for OPT_HOST_IFACE

drivers/block/nbd.c

@@ -239,8 +239,8 @@ static void nbd_dev_remove(struct nbd_device *nbd)
 
 	if (disk) {
 		del_gendisk(disk);
-		blk_mq_free_tag_set(&nbd->tag_set);
 		blk_cleanup_disk(disk);
+		blk_mq_free_tag_set(&nbd->tag_set);
 	}
 
 	/*

drivers/block/paride/pd.c

@@ -1014,8 +1014,8 @@ static void __exit pd_exit(void)
 		if (p) {
 			disk->gd = NULL;
 			del_gendisk(p);
-			blk_mq_free_tag_set(&disk->tag_set);
 			blk_cleanup_disk(p);
+			blk_mq_free_tag_set(&disk->tag_set);
 			pi_release(disk->pi);
 		}
 	}

drivers/block/xen-blkfront.c

@@ -502,34 +502,21 @@ static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
 static int blkif_ioctl(struct block_device *bdev, fmode_t mode,
 		       unsigned command, unsigned long argument)
 {
-	struct blkfront_info *info = bdev->bd_disk->private_data;
 	int i;
 
-	dev_dbg(&info->xbdev->dev, "command: 0x%x, argument: 0x%lx\n",
-		command, (long)argument);
-
 	switch (command) {
 	case CDROMMULTISESSION:
-		dev_dbg(&info->xbdev->dev, "FIXME: support multisession CDs later\n");
 		for (i = 0; i < sizeof(struct cdrom_multisession); i++)
 			if (put_user(0, (char __user *)(argument + i)))
 				return -EFAULT;
 		return 0;
-
-	case CDROM_GET_CAPABILITY: {
-		struct gendisk *gd = info->gd;
-		if (gd->flags & GENHD_FL_CD)
+	case CDROM_GET_CAPABILITY:
+		if (bdev->bd_disk->flags & GENHD_FL_CD)
 			return 0;
 		return -EINVAL;
-	}
-
 	default:
-		/*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
-		  command);*/
-		return -EINVAL; /* same return as native Linux */
+		return -EINVAL;
 	}
-
-	return 0;
 }
 
 static unsigned long blkif_ring_get_request(struct blkfront_ring_info *rinfo,
@@ -1177,36 +1164,6 @@ static int xlvbd_alloc_gendisk(blkif_sector_t capacity,
 	return err;
 }
 
-static void xlvbd_release_gendisk(struct blkfront_info *info)
-{
-	unsigned int minor, nr_minors, i;
-	struct blkfront_ring_info *rinfo;
-
-	if (info->rq == NULL)
-		return;
-
-	/* No more blkif_request(). */
-	blk_mq_stop_hw_queues(info->rq);
-
-	for_each_rinfo(info, rinfo, i) {
-		/* No more gnttab callback work. */
-		gnttab_cancel_free_callback(&rinfo->callback);
-
-		/* Flush gnttab callback work. Must be done with no locks held. */
-		flush_work(&rinfo->work);
-	}
-
-	del_gendisk(info->gd);
-
-	minor = info->gd->first_minor;
-	nr_minors = info->gd->minors;
-	xlbd_release_minors(minor, nr_minors);
-
-	blk_cleanup_disk(info->gd);
-	info->gd = NULL;
-	blk_mq_free_tag_set(&info->tag_set);
-}
-
 /* Already hold rinfo->ring_lock. */
 static inline void kick_pending_request_queues_locked(struct blkfront_ring_info *rinfo)
 {
@@ -1756,12 +1713,6 @@ static int write_per_ring_nodes(struct xenbus_transaction xbt,
 	return err;
 }
 
-static void free_info(struct blkfront_info *info)
-{
-	list_del(&info->info_list);
-	kfree(info);
-}
-
 /* Common code used when first setting up, and when resuming. */
 static int talk_to_blkback(struct xenbus_device *dev,
 			   struct blkfront_info *info)
@@ -1880,13 +1831,6 @@ static int talk_to_blkback(struct xenbus_device *dev,
 		xenbus_dev_fatal(dev, err, "%s", message);
  destroy_blkring:
 	blkif_free(info, 0);
-
-	mutex_lock(&blkfront_mutex);
-	free_info(info);
-	mutex_unlock(&blkfront_mutex);
-
-	dev_set_drvdata(&dev->dev, NULL);
-
 	return err;
 }
 
@@ -2126,38 +2070,26 @@ static int blkfront_resume(struct xenbus_device *dev)
 static void blkfront_closing(struct blkfront_info *info)
 {
 	struct xenbus_device *xbdev = info->xbdev;
-	struct block_device *bdev = NULL;
-
-	mutex_lock(&info->mutex);
+	struct blkfront_ring_info *rinfo;
+	unsigned int i;
 
-	if (xbdev->state == XenbusStateClosing) {
-		mutex_unlock(&info->mutex);
+	if (xbdev->state == XenbusStateClosing)
 		return;
-	}
 
-	if (info->gd)
-		bdev = bdgrab(info->gd->part0);
+	/* No more blkif_request(). */
+	blk_mq_stop_hw_queues(info->rq);
+	blk_set_queue_dying(info->rq);
+	set_capacity(info->gd, 0);
 
-	mutex_unlock(&info->mutex);
+	for_each_rinfo(info, rinfo, i) {
+		/* No more gnttab callback work. */
+		gnttab_cancel_free_callback(&rinfo->callback);
 
-	if (!bdev) {
-		xenbus_frontend_closed(xbdev);
-		return;
+		/* Flush gnttab callback work. Must be done with no locks held. */
+		flush_work(&rinfo->work);
 	}
 
-	mutex_lock(&bdev->bd_disk->open_mutex);
-
-	if (bdev->bd_openers) {
-		xenbus_dev_error(xbdev, -EBUSY,
-				 "Device in use; refusing to close");
-		xenbus_switch_state(xbdev, XenbusStateClosing);
-	} else {
-		xlvbd_release_gendisk(info);
 	xenbus_frontend_closed(xbdev);
-	}
-
-	mutex_unlock(&bdev->bd_disk->open_mutex);
-	bdput(bdev);
 }
 
 static void blkfront_setup_discard(struct blkfront_info *info)
@@ -2472,7 +2404,6 @@ static void blkback_changed(struct xenbus_device *dev,
 			break;
 		fallthrough;
 	case XenbusStateClosing:
-		if (info)
 		blkfront_closing(info);
 		break;
 	}
@@ -2481,56 +2412,21 @@ static void blkback_changed(struct xenbus_device *dev,
 static int blkfront_remove(struct xenbus_device *xbdev)
 {
 	struct blkfront_info *info = dev_get_drvdata(&xbdev->dev);
-	struct block_device *bdev = NULL;
-	struct gendisk *disk;
 
 	dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename);
 
-	if (!info)
-		return 0;
-
-	blkif_free(info, 0);
-
-	mutex_lock(&info->mutex);
-
-	disk = info->gd;
-	if (disk)
-		bdev = bdgrab(disk->part0);
-
-	info->xbdev = NULL;
-	mutex_unlock(&info->mutex);
-
-	if (!bdev) {
-		mutex_lock(&blkfront_mutex);
-		free_info(info);
-		mutex_unlock(&blkfront_mutex);
-		return 0;
-	}
-
-	/*
-	 * The xbdev was removed before we reached the Closed
-	 * state. See if it's safe to remove the disk. If the bdev
-	 * isn't closed yet, we let release take care of it.
-	 */
-
-	mutex_lock(&disk->open_mutex);
-	info = disk->private_data;
-
-	dev_warn(disk_to_dev(disk),
-		 "%s was hot-unplugged, %d stale handles\n",
-		 xbdev->nodename, bdev->bd_openers);
+	del_gendisk(info->gd);
 
-	if (info && !bdev->bd_openers) {
-		xlvbd_release_gendisk(info);
-		disk->private_data = NULL;
 	mutex_lock(&blkfront_mutex);
-		free_info(info);
+	list_del(&info->info_list);
 	mutex_unlock(&blkfront_mutex);
-	}
 
-	mutex_unlock(&disk->open_mutex);
-	bdput(bdev);
+	blkif_free(info, 0);
+	xlbd_release_minors(info->gd->first_minor, info->gd->minors);
+	blk_cleanup_disk(info->gd);
+	blk_mq_free_tag_set(&info->tag_set);
 
+	kfree(info);
 	return 0;
 }
 
@@ -2541,77 +2437,9 @@ static int blkfront_is_ready(struct xenbus_device *dev)
 	return info->is_ready && info->xbdev;
 }
 
-static int blkif_open(struct block_device *bdev, fmode_t mode)
-{
-	struct gendisk *disk = bdev->bd_disk;
-	struct blkfront_info *info;
-	int err = 0;
-
-	mutex_lock(&blkfront_mutex);
-
-	info = disk->private_data;
-	if (!info) {
-		/* xbdev gone */
-		err = -ERESTARTSYS;
-		goto out;
-	}
-
-	mutex_lock(&info->mutex);
-
-	if (!info->gd)
-		/* xbdev is closed */
-		err = -ERESTARTSYS;
-
-	mutex_unlock(&info->mutex);
-
-out:
-	mutex_unlock(&blkfront_mutex);
-	return err;
-}
-
-static void blkif_release(struct gendisk *disk, fmode_t mode)
-{
-	struct blkfront_info *info = disk->private_data;
-	struct xenbus_device *xbdev;
-
-	mutex_lock(&blkfront_mutex);
-	if (disk->part0->bd_openers)
-		goto out_mutex;
-
-	/*
-	 * Check if we have been instructed to close. We will have
-	 * deferred this request, because the bdev was still open.
-	 */
-
-	mutex_lock(&info->mutex);
-	xbdev = info->xbdev;
-
-	if (xbdev && xbdev->state == XenbusStateClosing) {
-		/* pending switch to state closed */
-		dev_info(disk_to_dev(disk), "releasing disk\n");
-		xlvbd_release_gendisk(info);
-		xenbus_frontend_closed(info->xbdev);
- 	}
-
-	mutex_unlock(&info->mutex);
-
-	if (!xbdev) {
-		/* sudden device removal */
-		dev_info(disk_to_dev(disk), "releasing disk\n");
-		xlvbd_release_gendisk(info);
-		disk->private_data = NULL;
-		free_info(info);
-	}
-
-out_mutex:
-	mutex_unlock(&blkfront_mutex);
-}
-
 static const struct block_device_operations xlvbd_block_fops =
 {
 	.owner = THIS_MODULE,
-	.open = blkif_open,
-	.release = blkif_release,
 	.getgeo = blkif_getgeo,
 	.ioctl = blkif_ioctl,
 	.compat_ioctl = blkdev_compat_ptr_ioctl,

drivers/nvme/host/pci.c

@@ -1554,6 +1554,28 @@ static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid)
 	wmb(); /* ensure the first interrupt sees the initialization */
 }
 
+/*
+ * Try getting shutdown_lock while setting up IO queues.
+ */
+static int nvme_setup_io_queues_trylock(struct nvme_dev *dev)
+{
+	/*
+	 * Give up if the lock is being held by nvme_dev_disable.
+	 */
+	if (!mutex_trylock(&dev->shutdown_lock))
+		return -ENODEV;
+
+	/*
+	 * Controller is in wrong state, fail early.
+	 */
+	if (dev->ctrl.state != NVME_CTRL_CONNECTING) {
+		mutex_unlock(&dev->shutdown_lock);
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
 static int nvme_create_queue(struct nvme_queue *nvmeq, int qid, bool polled)
 {
 	struct nvme_dev *dev = nvmeq->dev;
@@ -1582,8 +1604,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid, bool polled)
 		goto release_cq;
 
 	nvmeq->cq_vector = vector;
-	nvme_init_queue(nvmeq, qid);
 
+	result = nvme_setup_io_queues_trylock(dev);
+	if (result)
+		return result;
+	nvme_init_queue(nvmeq, qid);
 	if (!polled) {
 		result = queue_request_irq(nvmeq);
 		if (result < 0)
@@ -1591,10 +1616,12 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid, bool polled)
 	}
 
 	set_bit(NVMEQ_ENABLED, &nvmeq->flags);
+	mutex_unlock(&dev->shutdown_lock);
 	return result;
 
 release_sq:
 	dev->online_queues--;
+	mutex_unlock(&dev->shutdown_lock);
 	adapter_delete_sq(dev, qid);
 release_cq:
 	adapter_delete_cq(dev, qid);
@@ -2167,7 +2194,18 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	if (nr_io_queues == 0)
 		return 0;
 
-	clear_bit(NVMEQ_ENABLED, &adminq->flags);
+	/*
+	 * Free IRQ resources as soon as NVMEQ_ENABLED bit transitions
+	 * from set to unset. If there is a window to it is truely freed,
+	 * pci_free_irq_vectors() jumping into this window will crash.
+	 * And take lock to avoid racing with pci_free_irq_vectors() in
+	 * nvme_dev_disable() path.
+	 */
+	result = nvme_setup_io_queues_trylock(dev);
+	if (result)
+		return result;
+	if (test_and_clear_bit(NVMEQ_ENABLED, &adminq->flags))
+		pci_free_irq(pdev, 0, adminq);
 
 	if (dev->cmb_use_sqes) {
 		result = nvme_cmb_qdepth(dev, nr_io_queues,
@@ -2183,13 +2221,16 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 		result = nvme_remap_bar(dev, size);
 		if (!result)
 			break;
-		if (!--nr_io_queues)
-			return -ENOMEM;
+		if (!--nr_io_queues) {
+			result = -ENOMEM;
+			goto out_unlock;
+		}
 	} while (1);
 	adminq->q_db = dev->dbs;
 
  retry:
 	/* Deregister the admin queue's interrupt */
+	if (test_and_clear_bit(NVMEQ_ENABLED, &adminq->flags))
 		pci_free_irq(pdev, 0, adminq);
 
 	/*
@@ -2199,8 +2240,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	pci_free_irq_vectors(pdev);
 
 	result = nvme_setup_irqs(dev, nr_io_queues);
-	if (result <= 0)
-		return -EIO;
+	if (result <= 0) {
+		result = -EIO;
+		goto out_unlock;
+	}
 
 	dev->num_vecs = result;
 	result = max(result - 1, 1);
@@ -2214,8 +2257,9 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 */
 	result = queue_request_irq(adminq);
 	if (result)
-		return result;
+		goto out_unlock;
 	set_bit(NVMEQ_ENABLED, &adminq->flags);
+	mutex_unlock(&dev->shutdown_lock);
 
 	result = nvme_create_io_queues(dev);
 	if (result || dev->online_queues < 2)
@@ -2224,6 +2268,9 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	if (dev->online_queues - 1 < dev->max_qid) {
 		nr_io_queues = dev->online_queues - 1;
 		nvme_disable_io_queues(dev);
+		result = nvme_setup_io_queues_trylock(dev);
+		if (result)
+			return result;
 		nvme_suspend_io_queues(dev);
 		goto retry;
 	}
@@ -2232,6 +2279,9 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 					dev->io_queues[HCTX_TYPE_READ],
 					dev->io_queues[HCTX_TYPE_POLL]);
 	return 0;
+out_unlock:
+	mutex_unlock(&dev->shutdown_lock);
+	return result;
 }
 
 static void nvme_del_queue_end(struct request *req, blk_status_t error)
@@ -2962,7 +3012,6 @@ static void nvme_remove(struct pci_dev *pdev)
 	if (!pci_device_is_present(pdev)) {
 		nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_DEAD);
 		nvme_dev_disable(dev, true);
-		nvme_dev_remove_admin(dev);
 	}
 
 	flush_work(&dev->ctrl.reset_work);

drivers/nvme/host/tcp.c

@@ -123,7 +123,6 @@ struct nvme_tcp_ctrl {
 	struct blk_mq_tag_set	admin_tag_set;
 	struct sockaddr_storage addr;
 	struct sockaddr_storage src_addr;
-	struct net_device	*ndev;
 	struct nvme_ctrl	ctrl;
 
 	struct work_struct	err_work;
@@ -2533,8 +2532,7 @@ static struct nvme_ctrl *nvme_tcp_create_ctrl(struct device *dev,
 	}
 
 	if (opts->mask & NVMF_OPT_HOST_IFACE) {
-		ctrl->ndev = dev_get_by_name(&init_net, opts->host_iface);
-		if (!ctrl->ndev) {
+		if (!__dev_get_by_name(&init_net, opts->host_iface)) {
 			pr_err("invalid interface passed: %s\n",
 			       opts->host_iface);
 			ret = -ENODEV;