* fully tear down the queue in del_gendisk
@ 2022-06-19 6:05 Christoph Hellwig
2022-06-19 6:05 ` [PATCH 1/6] mtip32xx: remove the device_status debugfs file Christoph Hellwig
` (7 more replies)
0 siblings, 8 replies; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
Hi Jens,
this series simplifies teardown for most block drivers. Right now they
also have to call blk_cleanup_queue after calling del_gendisk, making
the teardown process rather confusing.
Instead this series records if the request_queue is owned by the gendisk,
which is always the case except for scsi and dasd or queues without a
gendisk at all, and then does the entire teardown in del_gendisk.
Note that while intended or 5.20, this series is generated against the
block-5.19 branch as that contains fixes in this area that haven't
made it to the for-5.10/block branch yet.
Diffstat:
arch/m68k/emu/nfblock.c | 4
arch/um/drivers/ubd_kern.c | 4
arch/xtensa/platforms/iss/simdisk.c | 4
block/blk-core.c | 43 -----
block/blk-mq-debugfs.c | 8
block/blk-mq.c | 43 ++++-
block/blk-sysfs.c | 5
block/blk.h | 3
block/bsg-lib.c | 4
block/genhd.c | 38 +---
drivers/block/amiflop.c | 2
drivers/block/aoe/aoeblk.c | 2
drivers/block/aoe/aoedev.c | 2
drivers/block/ataflop.c | 5
drivers/block/brd.c | 4
drivers/block/drbd/drbd_main.c | 4
drivers/block/floppy.c | 6
drivers/block/loop.c | 3
drivers/block/mtip32xx/mtip32xx.c | 298 +++++-------------------------------
drivers/block/mtip32xx/mtip32xx.h | 5
drivers/block/n64cart.c | 2
drivers/block/nbd.c | 4
drivers/block/null_blk/main.c | 4
drivers/block/paride/pcd.c | 4
drivers/block/paride/pd.c | 4
drivers/block/paride/pf.c | 4
drivers/block/pktcdvd.c | 4
drivers/block/ps3disk.c | 4
drivers/block/ps3vram.c | 4
drivers/block/rbd.c | 2
drivers/block/rnbd/rnbd-clt.c | 6
drivers/block/sunvdc.c | 4
drivers/block/swim.c | 2
drivers/block/swim3.c | 2
drivers/block/sx8.c | 6
drivers/block/virtio_blk.c | 3
drivers/block/xen-blkfront.c | 4
drivers/block/z2ram.c | 3
drivers/block/zram/zram_drv.c | 4
drivers/cdrom/gdrom.c | 3
drivers/md/bcache/super.c | 2
drivers/md/dm.c | 2
drivers/md/md.c | 4
drivers/memstick/core/ms_block.c | 3
drivers/memstick/core/mspro_block.c | 3
drivers/mmc/core/block.c | 1
drivers/mmc/core/queue.c | 1
drivers/mtd/mtd_blkdevs.c | 4
drivers/mtd/ubi/block.c | 4
drivers/nvdimm/btt.c | 4
drivers/nvdimm/pmem.c | 4
drivers/nvme/host/apple.c | 2
drivers/nvme/host/core.c | 3
drivers/nvme/host/fc.c | 12 -
drivers/nvme/host/multipath.c | 2
drivers/nvme/host/pci.c | 2
drivers/nvme/host/rdma.c | 12 -
drivers/nvme/host/tcp.c | 12 -
drivers/nvme/target/loop.c | 12 -
drivers/s390/block/dasd.c | 2
drivers/s390/block/dasd_genhd.c | 4
drivers/s390/block/dcssblk.c | 8
drivers/s390/block/scm_blk.c | 4
drivers/scsi/scsi_lib.c | 6
drivers/scsi/scsi_sysfs.c | 2
drivers/scsi/sd.c | 4
drivers/scsi/sr.c | 4
drivers/ufs/core/ufshcd.c | 4
include/linux/blk-mq.h | 3
include/linux/blkdev.h | 7
70 files changed, 225 insertions(+), 472 deletions(-)
^ permalink raw reply [flat|nested] 25+ messages in thread
* [PATCH 1/6] mtip32xx: remove the device_status debugfs file
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-20 8:37 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 2/6] mtip32xx: fix device removal Christoph Hellwig
` (6 subsequent siblings)
7 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
This file is a huge mess that iterates over all devices and is in the
way of fixing the device removal in this driver, so remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/block/mtip32xx/mtip32xx.c | 141 +-----------------------------
drivers/block/mtip32xx/mtip32xx.h | 4 -
2 files changed, 1 insertion(+), 144 deletions(-)
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index 27386a572ba49..4151c80f5bfcc 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -94,17 +94,12 @@
/* Device instance number, incremented each time a device is probed. */
static int instance;
-static LIST_HEAD(online_list);
-static LIST_HEAD(removing_list);
-static DEFINE_SPINLOCK(dev_lock);
-
/*
* Global variable used to hold the major block device number
* allocated in mtip_init().
*/
static int mtip_major;
static struct dentry *dfs_parent;
-static struct dentry *dfs_device_status;
static u32 cpu_use[NR_CPUS];
@@ -2170,106 +2165,6 @@ static const struct attribute_group *mtip_disk_attr_groups[] = {
NULL,
};
-/* debugsfs entries */
-
-static ssize_t show_device_status(struct device_driver *drv, char *buf)
-{
- int size = 0;
- struct driver_data *dd, *tmp;
- unsigned long flags;
- char id_buf[42];
- u16 status = 0;
-
- spin_lock_irqsave(&dev_lock, flags);
- size += sprintf(&buf[size], "Devices Present:\n");
- list_for_each_entry_safe(dd, tmp, &online_list, online_list) {
- if (dd->pdev) {
- if (dd->port &&
- dd->port->identify &&
- dd->port->identify_valid) {
- strlcpy(id_buf,
- (char *) (dd->port->identify + 10), 21);
- status = *(dd->port->identify + 141);
- } else {
- memset(id_buf, 0, 42);
- status = 0;
- }
-
- if (dd->port &&
- test_bit(MTIP_PF_REBUILD_BIT, &dd->port->flags)) {
- size += sprintf(&buf[size],
- " device %s %s (ftl rebuild %d %%)\n",
- dev_name(&dd->pdev->dev),
- id_buf,
- status);
- } else {
- size += sprintf(&buf[size],
- " device %s %s\n",
- dev_name(&dd->pdev->dev),
- id_buf);
- }
- }
- }
-
- size += sprintf(&buf[size], "Devices Being Removed:\n");
- list_for_each_entry_safe(dd, tmp, &removing_list, remove_list) {
- if (dd->pdev) {
- if (dd->port &&
- dd->port->identify &&
- dd->port->identify_valid) {
- strlcpy(id_buf,
- (char *) (dd->port->identify+10), 21);
- status = *(dd->port->identify + 141);
- } else {
- memset(id_buf, 0, 42);
- status = 0;
- }
-
- if (dd->port &&
- test_bit(MTIP_PF_REBUILD_BIT, &dd->port->flags)) {
- size += sprintf(&buf[size],
- " device %s %s (ftl rebuild %d %%)\n",
- dev_name(&dd->pdev->dev),
- id_buf,
- status);
- } else {
- size += sprintf(&buf[size],
- " device %s %s\n",
- dev_name(&dd->pdev->dev),
- id_buf);
- }
- }
- }
- spin_unlock_irqrestore(&dev_lock, flags);
-
- return size;
-}
-
-static ssize_t mtip_hw_read_device_status(struct file *f, char __user *ubuf,
- size_t len, loff_t *offset)
-{
- int size = *offset;
- char *buf;
- int rv = 0;
-
- if (!len || *offset)
- return 0;
-
- buf = kzalloc(MTIP_DFS_MAX_BUF_SIZE, GFP_KERNEL);
- if (!buf)
- return -ENOMEM;
-
- size += show_device_status(NULL, buf);
-
- *offset = size <= len ? size : len;
- size = copy_to_user(ubuf, buf, *offset);
- if (size)
- rv = -EFAULT;
-
- kfree(buf);
- return rv ? rv : *offset;
-}
-
static ssize_t mtip_hw_read_registers(struct file *f, char __user *ubuf,
size_t len, loff_t *offset)
{
@@ -2363,13 +2258,6 @@ static ssize_t mtip_hw_read_flags(struct file *f, char __user *ubuf,
return rv ? rv : *offset;
}
-static const struct file_operations mtip_device_status_fops = {
- .owner = THIS_MODULE,
- .open = simple_open,
- .read = mtip_hw_read_device_status,
- .llseek = no_llseek,
-};
-
static const struct file_operations mtip_regs_fops = {
.owner = THIS_MODULE,
.open = simple_open,
@@ -3905,7 +3793,6 @@ static int mtip_pci_probe(struct pci_dev *pdev,
const struct cpumask *node_mask;
int cpu, i = 0, j = 0;
int my_node = NUMA_NO_NODE;
- unsigned long flags;
/* Allocate memory for this devices private data. */
my_node = pcibus_to_node(pdev->bus);
@@ -3952,9 +3839,6 @@ static int mtip_pci_probe(struct pci_dev *pdev,
dd->pdev = pdev;
dd->numa_node = my_node;
- INIT_LIST_HEAD(&dd->online_list);
- INIT_LIST_HEAD(&dd->remove_list);
-
memset(dd->workq_name, 0, 32);
snprintf(dd->workq_name, 31, "mtipq%d", dd->instance);
@@ -4047,11 +3931,6 @@ static int mtip_pci_probe(struct pci_dev *pdev,
else
rv = 0; /* device in rebuild state, return 0 from probe */
- /* Add to online list even if in ftl rebuild */
- spin_lock_irqsave(&dev_lock, flags);
- list_add(&dd->online_list, &online_list);
- spin_unlock_irqrestore(&dev_lock, flags);
-
goto done;
block_initialize_err:
@@ -4085,15 +3964,10 @@ static int mtip_pci_probe(struct pci_dev *pdev,
static void mtip_pci_remove(struct pci_dev *pdev)
{
struct driver_data *dd = pci_get_drvdata(pdev);
- unsigned long flags, to;
+ unsigned long to;
set_bit(MTIP_DDF_REMOVAL_BIT, &dd->dd_flag);
- spin_lock_irqsave(&dev_lock, flags);
- list_del_init(&dd->online_list);
- list_add(&dd->remove_list, &removing_list);
- spin_unlock_irqrestore(&dev_lock, flags);
-
mtip_check_surprise_removal(dd);
synchronize_irq(dd->pdev->irq);
@@ -4124,10 +3998,6 @@ static void mtip_pci_remove(struct pci_dev *pdev)
pci_disable_msi(pdev);
- spin_lock_irqsave(&dev_lock, flags);
- list_del_init(&dd->remove_list);
- spin_unlock_irqrestore(&dev_lock, flags);
-
kfree(dd);
pcim_iounmap_regions(pdev, 1 << MTIP_ABAR);
@@ -4250,15 +4120,6 @@ static int __init mtip_init(void)
pr_warn("Error creating debugfs parent\n");
dfs_parent = NULL;
}
- if (dfs_parent) {
- dfs_device_status = debugfs_create_file("device_status",
- 0444, dfs_parent, NULL,
- &mtip_device_status_fops);
- if (IS_ERR_OR_NULL(dfs_device_status)) {
- pr_err("Error creating device_status node\n");
- dfs_device_status = NULL;
- }
- }
/* Register our PCI operations. */
error = pci_register_driver(&mtip_pci_driver);
diff --git a/drivers/block/mtip32xx/mtip32xx.h b/drivers/block/mtip32xx/mtip32xx.h
index 6816beb45352b..a80419c57bbe4 100644
--- a/drivers/block/mtip32xx/mtip32xx.h
+++ b/drivers/block/mtip32xx/mtip32xx.h
@@ -462,10 +462,6 @@ struct driver_data {
int isr_binding;
- struct list_head online_list; /* linkage for online list */
-
- struct list_head remove_list; /* linkage for removing list */
-
int unal_qdepth; /* qdepth of unaligned IO queue */
};
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* [PATCH 2/6] mtip32xx: fix device removal
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
2022-06-19 6:05 ` [PATCH 1/6] mtip32xx: remove the device_status debugfs file Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-20 8:38 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 3/6] block: remove QUEUE_FLAG_DEAD Christoph Hellwig
` (5 subsequent siblings)
7 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
Use the proper helper to mark a surpise removal, remove the gendisk as
soon as possible when removing the device and implement the ->free_disk
callback to ensure the private data is alive as long as the gendisk has
references.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/block/mtip32xx/mtip32xx.c | 157 +++++++++---------------------
drivers/block/mtip32xx/mtip32xx.h | 1 -
2 files changed, 44 insertions(+), 114 deletions(-)
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index 4151c80f5bfcc..e7604b3bf8a75 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -141,11 +141,8 @@ static bool mtip_check_surprise_removal(struct driver_data *dd)
pci_read_config_word(dd->pdev, 0x00, &vendor_id);
if (vendor_id == 0xFFFF) {
dd->sr = true;
- if (dd->queue)
- blk_queue_flag_set(QUEUE_FLAG_DEAD, dd->queue);
- else
- dev_warn(&dd->pdev->dev,
- "%s: dd->queue is NULL\n", __func__);
+ if (dd->disk)
+ blk_mark_disk_dead(dd->disk);
return true; /* device removed */
}
@@ -3185,26 +3182,12 @@ static int mtip_block_getgeo(struct block_device *dev,
return 0;
}
-static int mtip_block_open(struct block_device *dev, fmode_t mode)
+static void mtip_block_free_disk(struct gendisk *disk)
{
- struct driver_data *dd;
-
- if (dev && dev->bd_disk) {
- dd = (struct driver_data *) dev->bd_disk->private_data;
-
- if (dd) {
- if (test_bit(MTIP_DDF_REMOVAL_BIT,
- &dd->dd_flag)) {
- return -ENODEV;
- }
- return 0;
- }
- }
- return -ENODEV;
-}
+ struct driver_data *dd = disk->private_data;
-static void mtip_block_release(struct gendisk *disk, fmode_t mode)
-{
+ ida_free(&rssd_index_ida, dd->index);
+ kfree(dd);
}
/*
@@ -3214,13 +3197,12 @@ static void mtip_block_release(struct gendisk *disk, fmode_t mode)
* layer.
*/
static const struct block_device_operations mtip_block_ops = {
- .open = mtip_block_open,
- .release = mtip_block_release,
.ioctl = mtip_block_ioctl,
#ifdef CONFIG_COMPAT
.compat_ioctl = mtip_block_compat_ioctl,
#endif
.getgeo = mtip_block_getgeo,
+ .free_disk = mtip_block_free_disk,
.owner = THIS_MODULE
};
@@ -3561,72 +3543,6 @@ static int mtip_block_initialize(struct driver_data *dd)
return rv;
}
-static bool mtip_no_dev_cleanup(struct request *rq, void *data, bool reserv)
-{
- struct mtip_cmd *cmd = blk_mq_rq_to_pdu(rq);
-
- cmd->status = BLK_STS_IOERR;
- blk_mq_complete_request(rq);
- return true;
-}
-
-/*
- * Block layer deinitialization function.
- *
- * Called by the PCI layer as each P320 device is removed.
- *
- * @dd Pointer to the driver data structure.
- *
- * return value
- * 0
- */
-static int mtip_block_remove(struct driver_data *dd)
-{
- mtip_hw_debugfs_exit(dd);
-
- if (dd->mtip_svc_handler) {
- set_bit(MTIP_PF_SVC_THD_STOP_BIT, &dd->port->flags);
- wake_up_interruptible(&dd->port->svc_wait);
- kthread_stop(dd->mtip_svc_handler);
- }
-
- if (!dd->sr) {
- /*
- * Explicitly wait here for IOs to quiesce,
- * as mtip_standby_drive usually won't wait for IOs.
- */
- if (!mtip_quiesce_io(dd->port, MTIP_QUIESCE_IO_TIMEOUT_MS))
- mtip_standby_drive(dd);
- }
- else
- dev_info(&dd->pdev->dev, "device %s surprise removal\n",
- dd->disk->disk_name);
-
- blk_freeze_queue_start(dd->queue);
- blk_mq_quiesce_queue(dd->queue);
- blk_mq_tagset_busy_iter(&dd->tags, mtip_no_dev_cleanup, dd);
- blk_mq_unquiesce_queue(dd->queue);
-
- if (dd->disk) {
- if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag))
- del_gendisk(dd->disk);
- if (dd->disk->queue) {
- blk_cleanup_queue(dd->queue);
- blk_mq_free_tag_set(&dd->tags);
- dd->queue = NULL;
- }
- put_disk(dd->disk);
- }
- dd->disk = NULL;
-
- ida_free(&rssd_index_ida, dd->index);
-
- /* De-initialize the protocol layer. */
- mtip_hw_exit(dd);
-
- return 0;
-}
-
/*
* Function called by the PCI layer when just before the
* machine shuts down.
@@ -3643,23 +3559,15 @@ static int mtip_block_shutdown(struct driver_data *dd)
{
mtip_hw_shutdown(dd);
- /* Delete our gendisk structure, and cleanup the blk queue. */
- if (dd->disk) {
- dev_info(&dd->pdev->dev,
- "Shutting down %s ...\n", dd->disk->disk_name);
+ dev_info(&dd->pdev->dev,
+ "Shutting down %s ...\n", dd->disk->disk_name);
- if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag))
- del_gendisk(dd->disk);
- if (dd->disk->queue) {
- blk_cleanup_queue(dd->queue);
- blk_mq_free_tag_set(&dd->tags);
- }
- put_disk(dd->disk);
- dd->disk = NULL;
- dd->queue = NULL;
- }
+ if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag))
+ del_gendisk(dd->disk);
- ida_free(&rssd_index_ida, dd->index);
+ blk_cleanup_queue(dd->queue);
+ blk_mq_free_tag_set(&dd->tags);
+ put_disk(dd->disk);
return 0;
}
@@ -3966,8 +3874,6 @@ static void mtip_pci_remove(struct pci_dev *pdev)
struct driver_data *dd = pci_get_drvdata(pdev);
unsigned long to;
- set_bit(MTIP_DDF_REMOVAL_BIT, &dd->dd_flag);
-
mtip_check_surprise_removal(dd);
synchronize_irq(dd->pdev->irq);
@@ -3983,11 +3889,36 @@ static void mtip_pci_remove(struct pci_dev *pdev)
"Completion workers still active!\n");
}
- blk_mark_disk_dead(dd->disk);
set_bit(MTIP_DDF_REMOVE_PENDING_BIT, &dd->dd_flag);
- /* Clean up the block layer. */
- mtip_block_remove(dd);
+ if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag))
+ del_gendisk(dd->disk);
+
+ mtip_hw_debugfs_exit(dd);
+
+ if (dd->mtip_svc_handler) {
+ set_bit(MTIP_PF_SVC_THD_STOP_BIT, &dd->port->flags);
+ wake_up_interruptible(&dd->port->svc_wait);
+ kthread_stop(dd->mtip_svc_handler);
+ }
+
+ if (!dd->sr) {
+ /*
+ * Explicitly wait here for IOs to quiesce,
+ * as mtip_standby_drive usually won't wait for IOs.
+ */
+ if (!mtip_quiesce_io(dd->port, MTIP_QUIESCE_IO_TIMEOUT_MS))
+ mtip_standby_drive(dd);
+ }
+ else
+ dev_info(&dd->pdev->dev, "device %s surprise removal\n",
+ dd->disk->disk_name);
+
+ blk_cleanup_queue(dd->queue);
+ blk_mq_free_tag_set(&dd->tags);
+
+ /* De-initialize the protocol layer. */
+ mtip_hw_exit(dd);
if (dd->isr_workq) {
destroy_workqueue(dd->isr_workq);
@@ -3998,10 +3929,10 @@ static void mtip_pci_remove(struct pci_dev *pdev)
pci_disable_msi(pdev);
- kfree(dd);
-
pcim_iounmap_regions(pdev, 1 << MTIP_ABAR);
pci_set_drvdata(pdev, NULL);
+
+ put_disk(dd->disk);
}
/*
diff --git a/drivers/block/mtip32xx/mtip32xx.h b/drivers/block/mtip32xx/mtip32xx.h
index a80419c57bbe4..f7328f19ac5c2 100644
--- a/drivers/block/mtip32xx/mtip32xx.h
+++ b/drivers/block/mtip32xx/mtip32xx.h
@@ -149,7 +149,6 @@ enum {
MTIP_DDF_RESUME_BIT = 6,
MTIP_DDF_INIT_DONE_BIT = 7,
MTIP_DDF_REBUILD_FAILED_BIT = 8,
- MTIP_DDF_REMOVAL_BIT = 9,
MTIP_DDF_STOP_IO = ((1 << MTIP_DDF_REMOVE_PENDING_BIT) |
(1 << MTIP_DDF_SEC_LOCK_BIT) |
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* [PATCH 3/6] block: remove QUEUE_FLAG_DEAD
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
2022-06-19 6:05 ` [PATCH 1/6] mtip32xx: remove the device_status debugfs file Christoph Hellwig
2022-06-19 6:05 ` [PATCH 2/6] mtip32xx: fix device removal Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-19 14:07 ` Bart Van Assche
2022-06-20 8:39 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue Christoph Hellwig
` (4 subsequent siblings)
7 siblings, 2 replies; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
Disallow setting the blk-mq state on any queue that is already dying as
setting the state even then is a bad idea, and remove the now unused
QUEUE_FLAG_DEAD flag.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
block/blk-core.c | 3 ---
block/blk-mq-debugfs.c | 8 +++-----
include/linux/blkdev.h | 2 --
3 files changed, 3 insertions(+), 10 deletions(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 27fb1357ad4b8..088332984cd1b 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -313,9 +313,6 @@ void blk_cleanup_queue(struct request_queue *q)
* after draining finished.
*/
blk_freeze_queue(q);
-
- blk_queue_flag_set(QUEUE_FLAG_DEAD, q);
-
blk_sync_queue(q);
if (queue_is_mq(q)) {
blk_mq_cancel_work_sync(q);
diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c
index 4d1ce9ef43187..b80fae7ab1d95 100644
--- a/block/blk-mq-debugfs.c
+++ b/block/blk-mq-debugfs.c
@@ -116,7 +116,6 @@ static const char *const blk_queue_flag_name[] = {
QUEUE_FLAG_NAME(NOXMERGES),
QUEUE_FLAG_NAME(ADD_RANDOM),
QUEUE_FLAG_NAME(SAME_FORCE),
- QUEUE_FLAG_NAME(DEAD),
QUEUE_FLAG_NAME(INIT_DONE),
QUEUE_FLAG_NAME(STABLE_WRITES),
QUEUE_FLAG_NAME(POLL),
@@ -151,11 +150,10 @@ static ssize_t queue_state_write(void *data, const char __user *buf,
char opbuf[16] = { }, *op;
/*
- * The "state" attribute is removed after blk_cleanup_queue() has called
- * blk_mq_free_queue(). Return if QUEUE_FLAG_DEAD has been set to avoid
- * triggering a use-after-free.
+ * The "state" attribute is removed when the queue is removed. Don't
+ * allow setting the state on a dying queue to avoid a use-after-free.
*/
- if (blk_queue_dead(q))
+ if (blk_queue_dying(q))
return -ENOENT;
if (count >= sizeof(opbuf)) {
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 73c886eba8e19..c77ed4dcf561b 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -560,7 +560,6 @@ struct request_queue {
#define QUEUE_FLAG_NOXMERGES 9 /* No extended merges */
#define QUEUE_FLAG_ADD_RANDOM 10 /* Contributes to random pool */
#define QUEUE_FLAG_SAME_FORCE 12 /* force complete on same CPU */
-#define QUEUE_FLAG_DEAD 13 /* queue tear-down finished */
#define QUEUE_FLAG_INIT_DONE 14 /* queue is initialized */
#define QUEUE_FLAG_STABLE_WRITES 15 /* don't modify blks until WB is done */
#define QUEUE_FLAG_POLL 16 /* IO polling enabled if set */
@@ -588,7 +587,6 @@ bool blk_queue_flag_test_and_set(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_has_srcu(q) test_bit(QUEUE_FLAG_HAS_SRCU, &(q)->queue_flags)
-#define blk_queue_dead(q) test_bit(QUEUE_FLAG_DEAD, &(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)
#define blk_queue_noxmerges(q) \
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
` (2 preceding siblings ...)
2022-06-19 6:05 ` [PATCH 3/6] block: remove QUEUE_FLAG_DEAD Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-20 8:40 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 5/6] block: simplify disk shutdown Christoph Hellwig
` (3 subsequent siblings)
7 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
These flags only apply to file system I/O, and all file system I/O is
already drained by del_gendisk and thus can't be in progress when
blk_cleanup_queue is called.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
block/blk-core.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 088332984cd1b..2f418606e3bd3 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -304,9 +304,6 @@ void blk_cleanup_queue(struct request_queue *q)
blk_queue_flag_set(QUEUE_FLAG_DYING, q);
blk_queue_start_drain(q);
- blk_queue_flag_set(QUEUE_FLAG_NOMERGES, q);
- blk_queue_flag_set(QUEUE_FLAG_NOXMERGES, q);
-
/*
* Drain all requests queued before DYING marking. Set DEAD flag to
* prevent that blk_mq_run_hw_queues() accesses the hardware queues
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* [PATCH 5/6] block: simplify disk shutdown
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
` (3 preceding siblings ...)
2022-06-19 6:05 ` [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-20 8:47 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 6/6] block: remove blk_cleanup_disk Christoph Hellwig
` (2 subsequent siblings)
7 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
Set the queue dying flag and call blk_mq_exit_queue from del_gendisk for
all disks that do not have separately allocated queues, and thus remove
the need to call blk_cleanup_queue for them.
Rename blk_cleanup_disk to blk_mq_destroy_queue to make it clear that
this function is intended only for separately allocated blk-mq queues.
This saves an extra queue freeze for devices without a separately
allocated queue.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
block/blk-core.c | 37 -------------------------
block/blk-mq.c | 43 +++++++++++++++++++++++++++--
block/blk-sysfs.c | 5 ----
block/blk.h | 3 ++
block/bsg-lib.c | 4 +--
block/genhd.c | 23 ++++++++-------
drivers/block/ataflop.c | 1 -
drivers/block/loop.c | 1 -
drivers/block/mtip32xx/mtip32xx.c | 2 --
drivers/block/rnbd/rnbd-clt.c | 2 +-
drivers/block/sx8.c | 4 +--
drivers/block/virtio_blk.c | 1 -
drivers/block/z2ram.c | 1 -
drivers/cdrom/gdrom.c | 1 -
drivers/memstick/core/ms_block.c | 1 -
drivers/memstick/core/mspro_block.c | 1 -
drivers/mmc/core/block.c | 1 -
drivers/mmc/core/queue.c | 1 -
drivers/nvme/host/apple.c | 2 +-
drivers/nvme/host/core.c | 1 -
drivers/nvme/host/fc.c | 12 ++++----
drivers/nvme/host/pci.c | 2 +-
drivers/nvme/host/rdma.c | 12 ++++----
drivers/nvme/host/tcp.c | 12 ++++----
drivers/nvme/target/loop.c | 12 ++++----
drivers/s390/block/dasd.c | 2 +-
drivers/s390/block/dasd_genhd.c | 4 +--
drivers/scsi/scsi_lib.c | 6 ++--
drivers/scsi/scsi_sysfs.c | 2 +-
drivers/scsi/sd.c | 4 +--
drivers/scsi/sr.c | 4 +--
drivers/ufs/core/ufshcd.c | 4 +--
include/linux/blk-mq.h | 3 ++
include/linux/blkdev.h | 4 +--
34 files changed, 105 insertions(+), 113 deletions(-)
diff --git a/block/blk-core.c b/block/blk-core.c
index 2f418606e3bd3..409719b852095 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -284,43 +284,6 @@ void blk_queue_start_drain(struct request_queue *q)
wake_up_all(&q->mq_freeze_wq);
}
-/**
- * blk_cleanup_queue - shutdown a request queue
- * @q: request queue to shutdown
- *
- * Mark @q DYING, drain all pending requests, mark @q DEAD, destroy and
- * put it. All future requests will be failed immediately with -ENODEV.
- *
- * Context: can sleep
- */
-void blk_cleanup_queue(struct request_queue *q)
-{
- /* cannot be called from atomic context */
- might_sleep();
-
- WARN_ON_ONCE(blk_queue_registered(q));
-
- /* mark @q DYING, no new request or merges will be allowed afterwards */
- blk_queue_flag_set(QUEUE_FLAG_DYING, q);
- blk_queue_start_drain(q);
-
- /*
- * Drain all requests queued before DYING marking. Set DEAD flag to
- * prevent that blk_mq_run_hw_queues() accesses the hardware queues
- * after draining finished.
- */
- blk_freeze_queue(q);
- blk_sync_queue(q);
- if (queue_is_mq(q)) {
- blk_mq_cancel_work_sync(q);
- blk_mq_exit_queue(q);
- }
-
- /* @q is and will stay empty, shutdown and put */
- blk_put_queue(q);
-}
-EXPORT_SYMBOL(blk_cleanup_queue);
-
/**
* blk_queue_enter() - try to increase q->q_usage_counter
* @q: request queue pointer
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 33145ba52c960..784b8f35edca4 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -3890,7 +3890,7 @@ static struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set,
q->queuedata = queuedata;
ret = blk_mq_init_allocated_queue(set, q);
if (ret) {
- blk_cleanup_queue(q);
+ blk_put_queue(q);
return ERR_PTR(ret);
}
return q;
@@ -3902,6 +3902,35 @@ struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *set)
}
EXPORT_SYMBOL(blk_mq_init_queue);
+/**
+ * blk_mq_destroy_queue - shutdown a request queue
+ * @q: request queue to shutdown
+ *
+ * This shuts down a request queue allocated by blk_mq_init_queue() and drops
+ * the initial reference. All future requests will failed with -ENODEV.
+ *
+ * Context: can sleep
+ */
+void blk_mq_destroy_queue(struct request_queue *q)
+{
+ WARN_ON_ONCE(!queue_is_mq(q));
+ WARN_ON_ONCE(blk_queue_registered(q));
+
+ might_sleep();
+
+ blk_queue_flag_set(QUEUE_FLAG_DYING, q);
+ blk_queue_start_drain(q);
+ blk_freeze_queue(q);
+
+ blk_sync_queue(q);
+ blk_mq_cancel_work_sync(q);
+ blk_mq_exit_queue(q);
+
+ /* @q is and will stay empty, shutdown and put */
+ blk_put_queue(q);
+}
+EXPORT_SYMBOL(blk_mq_destroy_queue);
+
struct gendisk *__blk_mq_alloc_disk(struct blk_mq_tag_set *set, void *queuedata,
struct lock_class_key *lkclass)
{
@@ -3914,13 +3943,23 @@ struct gendisk *__blk_mq_alloc_disk(struct blk_mq_tag_set *set, void *queuedata,
disk = __alloc_disk_node(q, set->numa_node, lkclass);
if (!disk) {
- blk_cleanup_queue(q);
+ blk_put_queue(q);
return ERR_PTR(-ENOMEM);
}
+ set_bit(GD_OWNS_QUEUE, &disk->state);
return disk;
}
EXPORT_SYMBOL(__blk_mq_alloc_disk);
+struct gendisk *blk_mq_alloc_disk_for_queue(struct request_queue *q,
+ struct lock_class_key *lkclass)
+{
+ if (!blk_get_queue(q))
+ return NULL;
+ return __alloc_disk_node(q, NUMA_NO_NODE, lkclass);
+}
+EXPORT_SYMBOL(blk_mq_alloc_disk_for_queue);
+
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)
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index 9b905e9443e49..84d7f87015673 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -748,11 +748,6 @@ static void blk_free_queue_rcu(struct rcu_head *rcu_head)
* decremented with blk_put_queue(). Once the refcount reaches 0 this function
* is called.
*
- * For drivers that have a request_queue on a gendisk and added with
- * __device_add_disk() the refcount to request_queue will reach 0 with
- * the last put_disk() called by the driver. For drivers which don't use
- * __device_add_disk() this happens with blk_cleanup_queue().
- *
* Drivers exist which depend on the release of the request_queue to be
* synchronous, it should not be deferred.
*
diff --git a/block/blk.h b/block/blk.h
index 434017701403f..0d6668663ab5d 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -411,6 +411,9 @@ int bdev_resize_partition(struct gendisk *disk, int partno, sector_t start,
sector_t length);
void blk_drop_partitions(struct gendisk *disk);
+struct gendisk *__alloc_disk_node(struct request_queue *q, int node_id,
+ struct lock_class_key *lkclass);
+
int bio_add_hw_page(struct request_queue *q, struct bio *bio,
struct page *page, unsigned int len, unsigned int offset,
unsigned int max_sectors, bool *same_page);
diff --git a/block/bsg-lib.c b/block/bsg-lib.c
index acfe1357bf6c4..fd4cd5e682826 100644
--- a/block/bsg-lib.c
+++ b/block/bsg-lib.c
@@ -324,7 +324,7 @@ void bsg_remove_queue(struct request_queue *q)
container_of(q->tag_set, struct bsg_set, tag_set);
bsg_unregister_queue(bset->bd);
- blk_cleanup_queue(q);
+ blk_mq_destroy_queue(q);
blk_mq_free_tag_set(&bset->tag_set);
kfree(bset);
}
@@ -399,7 +399,7 @@ struct request_queue *bsg_setup_queue(struct device *dev, const char *name,
return q;
out_cleanup_queue:
- blk_cleanup_queue(q);
+ blk_mq_destroy_queue(q);
out_queue:
blk_mq_free_tag_set(set);
out_tag_set:
diff --git a/block/genhd.c b/block/genhd.c
index 278227ba1d531..4d15f828c4498 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -617,6 +617,8 @@ void del_gendisk(struct gendisk *disk)
* Fail any new I/O.
*/
set_bit(GD_DEAD, &disk->state);
+ if (test_bit(GD_OWNS_QUEUE, &disk->state))
+ blk_queue_flag_set(QUEUE_FLAG_DYING, q);
set_capacity(disk, 0);
/*
@@ -663,11 +665,16 @@ void del_gendisk(struct gendisk *disk)
blk_mq_unquiesce_queue(q);
/*
- * Allow using passthrough request again after the queue is torn down.
+ * If the disk does not own the queue, allow using passthrough requests
+ * again. Else leave the queue frozen to fail all I/O.
*/
- blk_queue_flag_clear(QUEUE_FLAG_INIT_DONE, q);
- __blk_mq_unfreeze_queue(q, true);
-
+ if (!test_bit(GD_OWNS_QUEUE, &disk->state)) {
+ blk_queue_flag_clear(QUEUE_FLAG_INIT_DONE, q);
+ __blk_mq_unfreeze_queue(q, true);
+ } else {
+ if (queue_is_mq(q))
+ blk_mq_exit_queue(q);
+ }
}
EXPORT_SYMBOL(del_gendisk);
@@ -1338,9 +1345,6 @@ struct gendisk *__alloc_disk_node(struct request_queue *q, int node_id,
{
struct gendisk *disk;
- if (!blk_get_queue(q))
- return NULL;
-
disk = kzalloc_node(sizeof(struct gendisk), GFP_KERNEL, node_id);
if (!disk)
goto out_put_queue;
@@ -1391,7 +1395,6 @@ struct gendisk *__alloc_disk_node(struct request_queue *q, int node_id,
blk_put_queue(q);
return NULL;
}
-EXPORT_SYMBOL(__alloc_disk_node);
struct gendisk *__blk_alloc_disk(int node, struct lock_class_key *lkclass)
{
@@ -1404,9 +1407,10 @@ struct gendisk *__blk_alloc_disk(int node, struct lock_class_key *lkclass)
disk = __alloc_disk_node(q, node, lkclass);
if (!disk) {
- blk_cleanup_queue(q);
+ blk_put_queue(q);
return NULL;
}
+ set_bit(GD_OWNS_QUEUE, &disk->state);
return disk;
}
EXPORT_SYMBOL(__blk_alloc_disk);
@@ -1439,7 +1443,6 @@ EXPORT_SYMBOL(put_disk);
*/
void blk_cleanup_disk(struct gendisk *disk)
{
- blk_cleanup_queue(disk->queue);
put_disk(disk);
}
EXPORT_SYMBOL(blk_cleanup_disk);
diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c
index e232cc4fd444b..c6e41ee18aaa2 100644
--- a/drivers/block/ataflop.c
+++ b/drivers/block/ataflop.c
@@ -2045,7 +2045,6 @@ static void atari_floppy_cleanup(void)
if (!unit[i].disk[type])
continue;
del_gendisk(unit[i].disk[type]);
- blk_cleanup_queue(unit[i].disk[type]->queue);
put_disk(unit[i].disk[type]);
}
blk_mq_free_tag_set(&unit[i].tag_set);
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 084f9b8a0ba3c..cc608226c8c78 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -2057,7 +2057,6 @@ static void loop_remove(struct loop_device *lo)
{
/* Make this loop device unreachable from pathname. */
del_gendisk(lo->lo_disk);
- blk_cleanup_queue(lo->lo_disk->queue);
blk_mq_free_tag_set(&lo->tag_set);
mutex_lock(&loop_ctl_mutex);
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index e7604b3bf8a75..1d0e0a9fdd7c2 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -3565,7 +3565,6 @@ static int mtip_block_shutdown(struct driver_data *dd)
if (test_bit(MTIP_DDF_INIT_DONE_BIT, &dd->dd_flag))
del_gendisk(dd->disk);
- blk_cleanup_queue(dd->queue);
blk_mq_free_tag_set(&dd->tags);
put_disk(dd->disk);
return 0;
@@ -3914,7 +3913,6 @@ static void mtip_pci_remove(struct pci_dev *pdev)
dev_info(&dd->pdev->dev, "device %s surprise removal\n",
dd->disk->disk_name);
- blk_cleanup_queue(dd->queue);
blk_mq_free_tag_set(&dd->tags);
/* De-initialize the protocol layer. */
diff --git a/drivers/block/rnbd/rnbd-clt.c b/drivers/block/rnbd/rnbd-clt.c
index 409c76b81aed4..a4470374f54fc 100644
--- a/drivers/block/rnbd/rnbd-clt.c
+++ b/drivers/block/rnbd/rnbd-clt.c
@@ -1755,7 +1755,7 @@ static void rnbd_destroy_sessions(void)
list_for_each_entry_safe(dev, tn, &sess->devs_list, list) {
/*
* Here unmap happens in parallel for only one reason:
- * blk_cleanup_queue() takes around half a second, so
+ * del_gendisk() takes around half a second, so
* on huge amount of devices the whole module unload
* procedure takes minutes.
*/
diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c
index 63b4f6431d2e6..75057dbbcfbea 100644
--- a/drivers/block/sx8.c
+++ b/drivers/block/sx8.c
@@ -1536,7 +1536,7 @@ static int carm_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
clear_bit(0, &carm_major_alloc);
else if (host->major == 161)
clear_bit(1, &carm_major_alloc);
- blk_cleanup_queue(host->oob_q);
+ blk_mq_destroy_queue(host->oob_q);
blk_mq_free_tag_set(&host->tag_set);
err_out_dma_free:
dma_free_coherent(&pdev->dev, CARM_SHM_SIZE, host->shm, host->shm_dma);
@@ -1570,7 +1570,7 @@ static void carm_remove_one (struct pci_dev *pdev)
clear_bit(0, &carm_major_alloc);
else if (host->major == 161)
clear_bit(1, &carm_major_alloc);
- blk_cleanup_queue(host->oob_q);
+ blk_mq_destroy_queue(host->oob_q);
blk_mq_free_tag_set(&host->tag_set);
dma_free_coherent(&pdev->dev, CARM_SHM_SIZE, host->shm, host->shm_dma);
iounmap(host->mmio);
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index 6fc7850c2b0a0..cff1b6f6b0548 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -1111,7 +1111,6 @@ static void virtblk_remove(struct virtio_device *vdev)
flush_work(&vblk->config_work);
del_gendisk(vblk->disk);
- blk_cleanup_queue(vblk->disk->queue);
blk_mq_free_tag_set(&vblk->tag_set);
mutex_lock(&vblk->vdev_mutex);
diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c
index 7a6ed83481b8d..18ad43d9933ec 100644
--- a/drivers/block/z2ram.c
+++ b/drivers/block/z2ram.c
@@ -384,7 +384,6 @@ static void __exit z2_exit(void)
for (i = 0; i < Z2MINOR_COUNT; i++) {
del_gendisk(z2ram_gendisk[i]);
- blk_cleanup_queue(z2ram_gendisk[i]->queue);
put_disk(z2ram_gendisk[i]);
}
blk_mq_free_tag_set(&tag_set);
diff --git a/drivers/cdrom/gdrom.c b/drivers/cdrom/gdrom.c
index 8e78b37d0f6a4..f4cc90ea6198e 100644
--- a/drivers/cdrom/gdrom.c
+++ b/drivers/cdrom/gdrom.c
@@ -831,7 +831,6 @@ static int probe_gdrom(struct platform_device *devptr)
static int remove_gdrom(struct platform_device *devptr)
{
- blk_cleanup_queue(gd.gdrom_rq);
blk_mq_free_tag_set(&gd.tag_set);
free_irq(HW_EVENT_GDROM_CMD, &gd);
free_irq(HW_EVENT_GDROM_DMA, &gd);
diff --git a/drivers/memstick/core/ms_block.c b/drivers/memstick/core/ms_block.c
index 3993bdd4b519c..ba7e7249a3db4 100644
--- a/drivers/memstick/core/ms_block.c
+++ b/drivers/memstick/core/ms_block.c
@@ -2187,7 +2187,6 @@ static void msb_remove(struct memstick_dev *card)
/* Remove the disk */
del_gendisk(msb->disk);
- blk_cleanup_queue(msb->queue);
blk_mq_free_tag_set(&msb->tag_set);
msb->queue = NULL;
diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c
index 725ba74ded308..72e91c06c618b 100644
--- a/drivers/memstick/core/mspro_block.c
+++ b/drivers/memstick/core/mspro_block.c
@@ -1294,7 +1294,6 @@ static void mspro_block_remove(struct memstick_dev *card)
del_gendisk(msb->disk);
dev_dbg(&card->dev, "mspro block remove\n");
- blk_cleanup_queue(msb->queue);
blk_mq_free_tag_set(&msb->tag_set);
msb->queue = NULL;
diff --git a/drivers/mmc/core/block.c b/drivers/mmc/core/block.c
index f4a1281658db0..bda6c67ce93f4 100644
--- a/drivers/mmc/core/block.c
+++ b/drivers/mmc/core/block.c
@@ -2509,7 +2509,6 @@ static struct mmc_blk_data *mmc_blk_alloc_req(struct mmc_card *card,
return md;
err_cleanup_queue:
- blk_cleanup_queue(md->disk->queue);
blk_mq_free_tag_set(&md->queue.tag_set);
err_kfree:
kfree(md);
diff --git a/drivers/mmc/core/queue.c b/drivers/mmc/core/queue.c
index fa5324ceeebe4..f824cfdab75ac 100644
--- a/drivers/mmc/core/queue.c
+++ b/drivers/mmc/core/queue.c
@@ -494,7 +494,6 @@ void mmc_cleanup_queue(struct mmc_queue *mq)
if (blk_queue_quiesced(q))
blk_mq_unquiesce_queue(q);
- blk_cleanup_queue(q);
blk_mq_free_tag_set(&mq->tag_set);
/*
diff --git a/drivers/nvme/host/apple.c b/drivers/nvme/host/apple.c
index d702d7d60235d..2d23b7d41f7e6 100644
--- a/drivers/nvme/host/apple.c
+++ b/drivers/nvme/host/apple.c
@@ -1502,7 +1502,7 @@ static int apple_nvme_probe(struct platform_device *pdev)
if (!blk_get_queue(anv->ctrl.admin_q)) {
nvme_start_admin_queue(&anv->ctrl);
- blk_cleanup_queue(anv->ctrl.admin_q);
+ blk_mq_destroy_queue(anv->ctrl.admin_q);
anv->ctrl.admin_q = NULL;
ret = -ENODEV;
goto put_dev;
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 3ab2cfd254a44..697af049b49d4 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -4089,7 +4089,6 @@ static void nvme_ns_remove(struct nvme_ns *ns)
if (!nvme_ns_head_multipath(ns->head))
nvme_cdev_del(&ns->cdev, &ns->cdev_device);
del_gendisk(ns->disk);
- blk_cleanup_queue(ns->queue);
down_write(&ns->ctrl->namespaces_rwsem);
list_del_init(&ns->list);
diff --git a/drivers/nvme/host/fc.c b/drivers/nvme/host/fc.c
index 3c778bb0c2944..a96aa831684c3 100644
--- a/drivers/nvme/host/fc.c
+++ b/drivers/nvme/host/fc.c
@@ -2392,7 +2392,7 @@ nvme_fc_ctrl_free(struct kref *ref)
unsigned long flags;
if (ctrl->ctrl.tagset) {
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
blk_mq_free_tag_set(&ctrl->tag_set);
}
@@ -2402,8 +2402,8 @@ nvme_fc_ctrl_free(struct kref *ref)
spin_unlock_irqrestore(&ctrl->rport->lock, flags);
nvme_start_admin_queue(&ctrl->ctrl);
- blk_cleanup_queue(ctrl->ctrl.admin_q);
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
blk_mq_free_tag_set(&ctrl->admin_tag_set);
kfree(ctrl->queues);
@@ -2953,7 +2953,7 @@ nvme_fc_create_io_queues(struct nvme_fc_ctrl *ctrl)
out_delete_hw_queues:
nvme_fc_delete_hw_io_queues(ctrl);
out_cleanup_blk_queue:
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
out_free_tag_set:
blk_mq_free_tag_set(&ctrl->tag_set);
nvme_fc_free_io_queues(ctrl);
@@ -3642,9 +3642,9 @@ nvme_fc_init_ctrl(struct device *dev, struct nvmf_ctrl_options *opts,
return ERR_PTR(-EIO);
out_cleanup_admin_q:
- blk_cleanup_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
out_cleanup_fabrics_q:
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
out_free_admin_tag_set:
blk_mq_free_tag_set(&ctrl->admin_tag_set);
out_free_queues:
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index c7012e85d035d..69c8b47b52274 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1760,7 +1760,7 @@ static void nvme_dev_remove_admin(struct nvme_dev *dev)
* queue to flush these to completion.
*/
nvme_start_admin_queue(&dev->ctrl);
- blk_cleanup_queue(dev->ctrl.admin_q);
+ blk_mq_destroy_queue(dev->ctrl.admin_q);
blk_mq_free_tag_set(&dev->admin_tagset);
}
}
diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index f2a5e1ea508a7..0fb7c8e7ab0be 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -840,8 +840,8 @@ static void nvme_rdma_destroy_admin_queue(struct nvme_rdma_ctrl *ctrl,
bool remove)
{
if (remove) {
- blk_cleanup_queue(ctrl->ctrl.admin_q);
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
blk_mq_free_tag_set(ctrl->ctrl.admin_tagset);
}
if (ctrl->async_event_sqe.data) {
@@ -935,10 +935,10 @@ static int nvme_rdma_configure_admin_queue(struct nvme_rdma_ctrl *ctrl,
nvme_cancel_admin_tagset(&ctrl->ctrl);
out_cleanup_queue:
if (new)
- blk_cleanup_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
out_cleanup_fabrics_q:
if (new)
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
out_free_tagset:
if (new)
blk_mq_free_tag_set(ctrl->ctrl.admin_tagset);
@@ -957,7 +957,7 @@ static void nvme_rdma_destroy_io_queues(struct nvme_rdma_ctrl *ctrl,
bool remove)
{
if (remove) {
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
blk_mq_free_tag_set(ctrl->ctrl.tagset);
}
nvme_rdma_free_io_queues(ctrl);
@@ -1012,7 +1012,7 @@ static int nvme_rdma_configure_io_queues(struct nvme_rdma_ctrl *ctrl, bool new)
out_cleanup_connect_q:
nvme_cancel_tagset(&ctrl->ctrl);
if (new)
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
out_free_tag_set:
if (new)
blk_mq_free_tag_set(ctrl->ctrl.tagset);
diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index bb67538d241b6..b81942fa5f959 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -1885,7 +1885,7 @@ static void nvme_tcp_destroy_io_queues(struct nvme_ctrl *ctrl, bool remove)
{
nvme_tcp_stop_io_queues(ctrl);
if (remove) {
- blk_cleanup_queue(ctrl->connect_q);
+ blk_mq_destroy_queue(ctrl->connect_q);
blk_mq_free_tag_set(ctrl->tagset);
}
nvme_tcp_free_io_queues(ctrl);
@@ -1940,7 +1940,7 @@ static int nvme_tcp_configure_io_queues(struct nvme_ctrl *ctrl, bool new)
out_cleanup_connect_q:
nvme_cancel_tagset(ctrl);
if (new)
- blk_cleanup_queue(ctrl->connect_q);
+ blk_mq_destroy_queue(ctrl->connect_q);
out_free_tag_set:
if (new)
blk_mq_free_tag_set(ctrl->tagset);
@@ -1953,8 +1953,8 @@ static void nvme_tcp_destroy_admin_queue(struct nvme_ctrl *ctrl, bool remove)
{
nvme_tcp_stop_queue(ctrl, 0);
if (remove) {
- blk_cleanup_queue(ctrl->admin_q);
- blk_cleanup_queue(ctrl->fabrics_q);
+ blk_mq_destroy_queue(ctrl->admin_q);
+ blk_mq_destroy_queue(ctrl->fabrics_q);
blk_mq_free_tag_set(ctrl->admin_tagset);
}
nvme_tcp_free_admin_queue(ctrl);
@@ -2012,10 +2012,10 @@ static int nvme_tcp_configure_admin_queue(struct nvme_ctrl *ctrl, bool new)
nvme_cancel_admin_tagset(ctrl);
out_cleanup_queue:
if (new)
- blk_cleanup_queue(ctrl->admin_q);
+ blk_mq_destroy_queue(ctrl->admin_q);
out_cleanup_fabrics_q:
if (new)
- blk_cleanup_queue(ctrl->fabrics_q);
+ blk_mq_destroy_queue(ctrl->fabrics_q);
out_free_tagset:
if (new)
blk_mq_free_tag_set(ctrl->admin_tagset);
diff --git a/drivers/nvme/target/loop.c b/drivers/nvme/target/loop.c
index 59024af2da2e3..0f5c77e22a0a9 100644
--- a/drivers/nvme/target/loop.c
+++ b/drivers/nvme/target/loop.c
@@ -266,8 +266,8 @@ static void nvme_loop_destroy_admin_queue(struct nvme_loop_ctrl *ctrl)
if (!test_and_clear_bit(NVME_LOOP_Q_LIVE, &ctrl->queues[0].flags))
return;
nvmet_sq_destroy(&ctrl->queues[0].nvme_sq);
- blk_cleanup_queue(ctrl->ctrl.admin_q);
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
blk_mq_free_tag_set(&ctrl->admin_tag_set);
}
@@ -283,7 +283,7 @@ static void nvme_loop_free_ctrl(struct nvme_ctrl *nctrl)
mutex_unlock(&nvme_loop_ctrl_mutex);
if (nctrl->tagset) {
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
blk_mq_free_tag_set(&ctrl->tag_set);
}
kfree(ctrl->queues);
@@ -410,9 +410,9 @@ static int nvme_loop_configure_admin_queue(struct nvme_loop_ctrl *ctrl)
out_cleanup_queue:
clear_bit(NVME_LOOP_Q_LIVE, &ctrl->queues[0].flags);
- blk_cleanup_queue(ctrl->ctrl.admin_q);
+ blk_mq_destroy_queue(ctrl->ctrl.admin_q);
out_cleanup_fabrics_q:
- blk_cleanup_queue(ctrl->ctrl.fabrics_q);
+ blk_mq_destroy_queue(ctrl->ctrl.fabrics_q);
out_free_tagset:
blk_mq_free_tag_set(&ctrl->admin_tag_set);
out_free_sq:
@@ -554,7 +554,7 @@ static int nvme_loop_create_io_queues(struct nvme_loop_ctrl *ctrl)
return 0;
out_cleanup_connect_q:
- blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_destroy_queue(ctrl->ctrl.connect_q);
out_free_tagset:
blk_mq_free_tag_set(&ctrl->tag_set);
out_destroy_queues:
diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c
index ba6d787896606..e8489331f12b8 100644
--- a/drivers/s390/block/dasd.c
+++ b/drivers/s390/block/dasd.c
@@ -3280,7 +3280,7 @@ static int dasd_alloc_queue(struct dasd_block *block)
static void dasd_free_queue(struct dasd_block *block)
{
if (block->request_queue) {
- blk_cleanup_queue(block->request_queue);
+ blk_mq_destroy_queue(block->request_queue);
blk_mq_free_tag_set(&block->tag_set);
block->request_queue = NULL;
}
diff --git a/drivers/s390/block/dasd_genhd.c b/drivers/s390/block/dasd_genhd.c
index a7a33ebf4bbe9..5a83f0a39901b 100644
--- a/drivers/s390/block/dasd_genhd.c
+++ b/drivers/s390/block/dasd_genhd.c
@@ -41,8 +41,8 @@ int dasd_gendisk_alloc(struct dasd_block *block)
if (base->devindex >= DASD_PER_MAJOR)
return -EBUSY;
- gdp = __alloc_disk_node(block->request_queue, NUMA_NO_NODE,
- &dasd_bio_compl_lkclass);
+ gdp = blk_mq_alloc_disk_for_queue(block->request_queue,
+ &dasd_bio_compl_lkclass);
if (!gdp)
return -ENOMEM;
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 6ffc9e4258a80..cdf0056582d5f 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -163,7 +163,7 @@ static void __scsi_queue_insert(struct scsi_cmnd *cmd, int reason, bool unbusy)
* Requeue this command. It will go before all other commands
* that are already in the queue. Schedule requeue work under
* lock such that the kblockd_schedule_work() call happens
- * before blk_cleanup_queue() finishes.
+ * before blk_mq_destroy_queue() finishes.
*/
cmd->result = 0;
@@ -424,9 +424,9 @@ static void scsi_starved_list_run(struct Scsi_Host *shost)
* it and the queue. Mitigate by taking a reference to the
* queue and never touching the sdev again after we drop the
* host lock. Note: if __scsi_remove_device() invokes
- * blk_cleanup_queue() before the queue is run from this
+ * blk_mq_destroy_queue() before the queue is run from this
* function then blk_run_queue() will return immediately since
- * blk_cleanup_queue() marks the queue with QUEUE_FLAG_DYING.
+ * blk_mq_destroy_queue() marks the queue with QUEUE_FLAG_DYING.
*/
slq = sdev->request_queue;
if (!blk_get_queue(slq))
diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c
index 43949798a2e47..aa70d9282161d 100644
--- a/drivers/scsi/scsi_sysfs.c
+++ b/drivers/scsi/scsi_sysfs.c
@@ -1475,7 +1475,7 @@ void __scsi_remove_device(struct scsi_device *sdev)
scsi_device_set_state(sdev, SDEV_DEL);
mutex_unlock(&sdev->state_mutex);
- blk_cleanup_queue(sdev->request_queue);
+ blk_mq_destroy_queue(sdev->request_queue);
cancel_work_sync(&sdev->requeue_work);
if (sdev->host->hostt->slave_destroy)
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index a1a2ac09066fd..cb587e488601c 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -3440,8 +3440,8 @@ static int sd_probe(struct device *dev)
if (!sdkp)
goto out;
- gd = __alloc_disk_node(sdp->request_queue, NUMA_NO_NODE,
- &sd_bio_compl_lkclass);
+ gd = blk_mq_alloc_disk_for_queue(sdp->request_queue,
+ &sd_bio_compl_lkclass);
if (!gd)
goto out_free;
diff --git a/drivers/scsi/sr.c b/drivers/scsi/sr.c
index 32d3b8274f148..a278b739d0c5f 100644
--- a/drivers/scsi/sr.c
+++ b/drivers/scsi/sr.c
@@ -624,8 +624,8 @@ static int sr_probe(struct device *dev)
if (!cd)
goto fail;
- disk = __alloc_disk_node(sdev->request_queue, NUMA_NO_NODE,
- &sr_bio_compl_lkclass);
+ disk = blk_mq_alloc_disk_for_queue(sdev->request_queue,
+ &sr_bio_compl_lkclass);
if (!disk)
goto fail_free;
mutex_init(&cd->lock);
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index 01fb4bad86be8..a46330f9f973a 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -9467,7 +9467,7 @@ void ufshcd_remove(struct ufs_hba *hba)
ufs_bsg_remove(hba);
ufshpb_remove(hba);
ufs_sysfs_remove_nodes(hba->dev);
- blk_cleanup_queue(hba->tmf_queue);
+ blk_mq_destroy_queue(hba->tmf_queue);
blk_mq_free_tag_set(&hba->tmf_tag_set);
scsi_remove_host(hba->host);
/* disable interrupts */
@@ -9763,7 +9763,7 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq)
return 0;
free_tmf_queue:
- blk_cleanup_queue(hba->tmf_queue);
+ blk_mq_destroy_queue(hba->tmf_queue);
free_tmf_tag_set:
blk_mq_free_tag_set(&hba->tmf_tag_set);
out_remove_scsi_host:
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index e2d9daf7e8dd0..0fd96e92c6c65 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -686,10 +686,13 @@ struct gendisk *__blk_mq_alloc_disk(struct blk_mq_tag_set *set, void *queuedata,
\
__blk_mq_alloc_disk(set, queuedata, &__key); \
})
+struct gendisk *blk_mq_alloc_disk_for_queue(struct request_queue *q,
+ struct lock_class_key *lkclass);
struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *);
int blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
struct request_queue *q);
void blk_mq_unregister_dev(struct device *, struct request_queue *);
+void blk_mq_destroy_queue(struct request_queue *);
int blk_mq_alloc_tag_set(struct blk_mq_tag_set *set);
int blk_mq_alloc_sq_tag_set(struct blk_mq_tag_set *set,
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index c77ed4dcf561b..b9bf092531bc7 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -148,6 +148,7 @@ struct gendisk {
#define GD_NATIVE_CAPACITY 3
#define GD_ADDED 4
#define GD_SUPPRESS_PART_SCAN 5
+#define GD_OWNS_QUEUE 6
struct mutex open_mutex; /* open/close mutex */
unsigned open_partitions; /* number of open partitions */
@@ -811,8 +812,6 @@ static inline u64 sb_bdev_nr_blocks(struct super_block *sb)
int bdev_disk_changed(struct gendisk *disk, bool invalidate);
-struct gendisk *__alloc_disk_node(struct request_queue *q, int node_id,
- struct lock_class_key *lkclass);
void put_disk(struct gendisk *disk);
struct gendisk *__blk_alloc_disk(int node, struct lock_class_key *lkclass);
@@ -954,7 +953,6 @@ static inline unsigned int blk_max_size_offset(struct request_queue *q,
/*
* Access functions for manipulating queue properties
*/
-extern void blk_cleanup_queue(struct request_queue *);
void blk_queue_bounce_limit(struct request_queue *q, enum blk_bounce limit);
extern void blk_queue_max_hw_sectors(struct request_queue *, unsigned int);
extern void blk_queue_chunk_sectors(struct request_queue *, unsigned int);
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* [PATCH 6/6] block: remove blk_cleanup_disk
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
` (4 preceding siblings ...)
2022-06-19 6:05 ` [PATCH 5/6] block: simplify disk shutdown Christoph Hellwig
@ 2022-06-19 6:05 ` Christoph Hellwig
2022-06-20 8:49 ` Hannes Reinecke
2022-06-19 22:21 ` fully tear down the queue in del_gendisk Jens Axboe
2022-06-28 12:33 ` Jens Axboe
7 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-19 6:05 UTC (permalink / raw)
To: Jens Axboe; +Cc: Ming Lei, linux-block
blk_cleanup_disk is nothing but a trivial wrapper for put_disk now,
so remove it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
arch/m68k/emu/nfblock.c | 4 ++--
arch/um/drivers/ubd_kern.c | 4 ++--
arch/xtensa/platforms/iss/simdisk.c | 4 ++--
block/genhd.c | 15 ---------------
drivers/block/amiflop.c | 2 +-
drivers/block/aoe/aoeblk.c | 2 +-
drivers/block/aoe/aoedev.c | 2 +-
drivers/block/ataflop.c | 4 ++--
drivers/block/brd.c | 4 ++--
drivers/block/drbd/drbd_main.c | 4 ++--
drivers/block/floppy.c | 6 +++---
drivers/block/loop.c | 2 +-
drivers/block/mtip32xx/mtip32xx.c | 2 +-
drivers/block/n64cart.c | 2 +-
drivers/block/nbd.c | 4 ++--
drivers/block/null_blk/main.c | 4 ++--
drivers/block/paride/pcd.c | 4 ++--
drivers/block/paride/pd.c | 4 ++--
drivers/block/paride/pf.c | 4 ++--
drivers/block/pktcdvd.c | 4 ++--
drivers/block/ps3disk.c | 4 ++--
drivers/block/ps3vram.c | 4 ++--
drivers/block/rbd.c | 2 +-
drivers/block/rnbd/rnbd-clt.c | 4 ++--
drivers/block/sunvdc.c | 4 ++--
drivers/block/swim.c | 2 +-
drivers/block/swim3.c | 2 +-
drivers/block/sx8.c | 2 +-
drivers/block/virtio_blk.c | 2 +-
drivers/block/xen-blkfront.c | 4 ++--
drivers/block/z2ram.c | 2 +-
drivers/block/zram/zram_drv.c | 4 ++--
drivers/cdrom/gdrom.c | 2 +-
drivers/md/bcache/super.c | 2 +-
drivers/md/dm.c | 2 +-
drivers/md/md.c | 4 ++--
drivers/memstick/core/ms_block.c | 2 +-
drivers/memstick/core/mspro_block.c | 2 +-
drivers/mtd/mtd_blkdevs.c | 4 ++--
drivers/mtd/ubi/block.c | 4 ++--
drivers/nvdimm/btt.c | 4 ++--
drivers/nvdimm/pmem.c | 4 ++--
drivers/nvme/host/core.c | 2 +-
drivers/nvme/host/multipath.c | 2 +-
drivers/s390/block/dcssblk.c | 8 ++++----
drivers/s390/block/scm_blk.c | 4 ++--
include/linux/blkdev.h | 1 -
47 files changed, 74 insertions(+), 90 deletions(-)
diff --git a/arch/m68k/emu/nfblock.c b/arch/m68k/emu/nfblock.c
index 267b02cc5655b..a708fbd5a844f 100644
--- a/arch/m68k/emu/nfblock.c
+++ b/arch/m68k/emu/nfblock.c
@@ -138,7 +138,7 @@ static int __init nfhd_init_one(int id, u32 blocks, u32 bsize)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(dev->disk);
+ put_disk(dev->disk);
free_dev:
kfree(dev);
out:
@@ -180,7 +180,7 @@ static void __exit nfhd_exit(void)
list_for_each_entry_safe(dev, next, &nfhd_list, list) {
list_del(&dev->list);
del_gendisk(dev->disk);
- blk_cleanup_disk(dev->disk);
+ put_disk(dev->disk);
kfree(dev);
}
unregister_blkdev(major_num, "nfhd");
diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c
index c4344b67628dd..479b79e114424 100644
--- a/arch/um/drivers/ubd_kern.c
+++ b/arch/um/drivers/ubd_kern.c
@@ -925,7 +925,7 @@ static int ubd_add(int n, char **error_out)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_cleanup_tags:
blk_mq_free_tag_set(&ubd_dev->tag_set);
out:
@@ -1032,7 +1032,7 @@ static int ubd_remove(int n, char **error_out)
ubd_gendisk[n] = NULL;
if(disk != NULL){
del_gendisk(disk);
- blk_cleanup_disk(disk);
+ put_disk(disk);
}
err = 0;
diff --git a/arch/xtensa/platforms/iss/simdisk.c b/arch/xtensa/platforms/iss/simdisk.c
index 4255b92fa3eb0..f50caaa1c2496 100644
--- a/arch/xtensa/platforms/iss/simdisk.c
+++ b/arch/xtensa/platforms/iss/simdisk.c
@@ -290,7 +290,7 @@ static int __init simdisk_setup(struct simdisk *dev, int which,
return 0;
out_cleanup_disk:
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
out:
return err;
}
@@ -344,7 +344,7 @@ static void simdisk_teardown(struct simdisk *dev, int which,
simdisk_detach(dev);
if (dev->gd) {
del_gendisk(dev->gd);
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
}
remove_proc_entry(tmp, procdir);
}
diff --git a/block/genhd.c b/block/genhd.c
index 4d15f828c4498..bf9be06af2c8d 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -1432,21 +1432,6 @@ void put_disk(struct gendisk *disk)
}
EXPORT_SYMBOL(put_disk);
-/**
- * blk_cleanup_disk - shutdown a gendisk allocated by blk_alloc_disk
- * @disk: gendisk to shutdown
- *
- * Mark the queue hanging off @disk DYING, drain all pending requests, then mark
- * the queue DEAD, destroy and put it and the gendisk structure.
- *
- * Context: can sleep
- */
-void blk_cleanup_disk(struct gendisk *disk)
-{
- put_disk(disk);
-}
-EXPORT_SYMBOL(blk_cleanup_disk);
-
static void set_disk_ro_uevent(struct gendisk *gd, int ro)
{
char event[] = "DISK_RO=1";
diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c
index 5a566f2fd533a..4c8b2ba579ee6 100644
--- a/drivers/block/amiflop.c
+++ b/drivers/block/amiflop.c
@@ -1802,7 +1802,7 @@ static int fd_alloc_disk(int drive, int system)
unit[drive].gendisk[system] = disk;
err = add_disk(disk);
if (err)
- blk_cleanup_disk(disk);
+ put_disk(disk);
return err;
}
diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c
index 348adf3352177..12b3ca8f6f4a9 100644
--- a/drivers/block/aoe/aoeblk.c
+++ b/drivers/block/aoe/aoeblk.c
@@ -427,7 +427,7 @@ aoeblk_gdalloc(void *vp)
return;
out_disk_cleanup:
- blk_cleanup_disk(gd);
+ put_disk(gd);
err_tagset:
blk_mq_free_tag_set(set);
err_mempool:
diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c
index b381d1c3ef327..3523dd82d7a00 100644
--- a/drivers/block/aoe/aoedev.c
+++ b/drivers/block/aoe/aoedev.c
@@ -277,7 +277,7 @@ freedev(struct aoedev *d)
if (d->gd) {
aoedisk_rm_debugfs(d);
del_gendisk(d->gd);
- blk_cleanup_disk(d->gd);
+ put_disk(d->gd);
blk_mq_free_tag_set(&d->tag_set);
}
t = d->targets;
diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c
index c6e41ee18aaa2..9deb4df6bdb8b 100644
--- a/drivers/block/ataflop.c
+++ b/drivers/block/ataflop.c
@@ -2031,7 +2031,7 @@ static void ataflop_probe(dev_t dev)
return;
cleanup_disk:
- blk_cleanup_disk(unit[drive].disk[type]);
+ put_disk(unit[drive].disk[type]);
unit[drive].disk[type] = NULL;
}
@@ -2063,7 +2063,7 @@ static void atari_cleanup_floppy_disk(struct atari_floppy_struct *fs)
continue;
if (fs->registered[type])
del_gendisk(fs->disk[type]);
- blk_cleanup_disk(fs->disk[type]);
+ put_disk(fs->disk[type]);
}
blk_mq_free_tag_set(&fs->tag_set);
}
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index 6e3f2f0d23520..9e26d5e769f32 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -419,7 +419,7 @@ static int brd_alloc(int i)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_free_dev:
list_del(&brd->brd_list);
kfree(brd);
@@ -439,7 +439,7 @@ static void brd_cleanup(void)
list_for_each_entry_safe(brd, next, &brd_devices, brd_list) {
del_gendisk(brd->brd_disk);
- blk_cleanup_disk(brd->brd_disk);
+ put_disk(brd->brd_disk);
brd_free_pages(brd);
list_del(&brd->brd_list);
kfree(brd);
diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index 2887350ae010a..f3e4db16fd07b 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -2207,7 +2207,7 @@ void drbd_destroy_device(struct kref *kref)
if (device->bitmap) /* should no longer be there. */
drbd_bm_cleanup(device);
__free_page(device->md_io.page);
- blk_cleanup_disk(device->vdisk);
+ put_disk(device->vdisk);
kfree(device->rs_plan_s);
/* not for_each_connection(connection, resource):
@@ -2807,7 +2807,7 @@ enum drbd_ret_code drbd_create_device(struct drbd_config_context *adm_ctx, unsig
out_no_bitmap:
__free_page(device->md_io.page);
out_no_io_page:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_no_disk:
kref_put(&resource->kref, drbd_destroy_resource);
kfree(device);
diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c
index 015841f50f4e9..491e7205a0db0 100644
--- a/drivers/block/floppy.c
+++ b/drivers/block/floppy.c
@@ -4557,7 +4557,7 @@ static void floppy_probe(dev_t dev)
return;
cleanup_disk:
- blk_cleanup_disk(disks[drive][type]);
+ put_disk(disks[drive][type]);
disks[drive][type] = NULL;
mutex_unlock(&floppy_probe_lock);
}
@@ -4753,7 +4753,7 @@ static int __init do_floppy_init(void)
if (!disks[drive][0])
break;
del_timer_sync(&motor_off_timer[drive]);
- blk_cleanup_disk(disks[drive][0]);
+ put_disk(disks[drive][0]);
blk_mq_free_tag_set(&tag_sets[drive]);
}
return err;
@@ -4985,7 +4985,7 @@ static void __exit floppy_module_exit(void)
}
for (i = 0; i < ARRAY_SIZE(floppy_type); i++) {
if (disks[drive][i])
- blk_cleanup_disk(disks[drive][i]);
+ put_disk(disks[drive][i]);
}
blk_mq_free_tag_set(&tag_sets[drive]);
}
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index cc608226c8c78..e3c0ba93c1a34 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -2040,7 +2040,7 @@ static int loop_add(int i)
return i;
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_cleanup_tags:
blk_mq_free_tag_set(&lo->tag_set);
out_free_idr:
diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c
index 1d0e0a9fdd7c2..e116c6cf56f5f 100644
--- a/drivers/block/mtip32xx/mtip32xx.c
+++ b/drivers/block/mtip32xx/mtip32xx.c
@@ -3534,7 +3534,7 @@ static int mtip_block_initialize(struct driver_data *dd)
disk_index_error:
ida_free(&rssd_index_ida, index);
ida_get_error:
- blk_cleanup_disk(dd->disk);
+ put_disk(dd->disk);
block_queue_alloc_init_error:
blk_mq_free_tag_set(&dd->tags);
block_queue_alloc_tag_error:
diff --git a/drivers/block/n64cart.c b/drivers/block/n64cart.c
index e094d2b8b5a92..d914156db2d8b 100644
--- a/drivers/block/n64cart.c
+++ b/drivers/block/n64cart.c
@@ -157,7 +157,7 @@ static int __init n64cart_probe(struct platform_device *pdev)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out:
return err;
}
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 07f3c139a3d77..5c4c9c45c6ace 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -250,7 +250,7 @@ static void nbd_dev_remove(struct nbd_device *nbd)
struct gendisk *disk = nbd->disk;
del_gendisk(disk);
- blk_cleanup_disk(disk);
+ put_disk(disk);
blk_mq_free_tag_set(&nbd->tag_set);
/*
@@ -1833,7 +1833,7 @@ static struct nbd_device *nbd_dev_add(int index, unsigned int refs)
out_free_work:
destroy_workqueue(nbd->recv_workq);
out_err_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_free_idr:
mutex_lock(&nbd_index_mutex);
idr_remove(&nbd_index_idr, index);
diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c
index 6b67088f4ea71..d695ea29efa6d 100644
--- a/drivers/block/null_blk/main.c
+++ b/drivers/block/null_blk/main.c
@@ -1737,7 +1737,7 @@ static void null_del_dev(struct nullb *nullb)
null_restart_queue_async(nullb);
}
- blk_cleanup_disk(nullb->disk);
+ put_disk(nullb->disk);
if (dev->queue_mode == NULL_Q_MQ &&
nullb->tag_set == &nullb->__tag_set)
blk_mq_free_tag_set(nullb->tag_set);
@@ -2082,7 +2082,7 @@ static int null_add_dev(struct nullb_device *dev)
out_cleanup_zone:
null_free_zoned_dev(dev);
out_cleanup_disk:
- blk_cleanup_disk(nullb->disk);
+ put_disk(nullb->disk);
out_cleanup_tags:
if (dev->queue_mode == NULL_Q_MQ && nullb->tag_set == &nullb->__tag_set)
blk_mq_free_tag_set(nullb->tag_set);
diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c
index f462ad67931aa..a5ab407841193 100644
--- a/drivers/block/paride/pcd.c
+++ b/drivers/block/paride/pcd.c
@@ -956,7 +956,7 @@ static int pcd_init_unit(struct pcd_unit *cd, bool autoprobe, int port,
out_pi_release:
pi_release(cd->pi);
out_free_disk:
- blk_cleanup_disk(cd->disk);
+ put_disk(cd->disk);
out_free_tag_set:
blk_mq_free_tag_set(&cd->tag_set);
return ret;
@@ -1029,7 +1029,7 @@ static void __exit pcd_exit(void)
unregister_cdrom(&cd->info);
del_gendisk(cd->disk);
pi_release(cd->pi);
- blk_cleanup_disk(cd->disk);
+ put_disk(cd->disk);
blk_mq_free_tag_set(&cd->tag_set);
}
diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c
index 3637c38c72f97..c8c14c6f5c3a2 100644
--- a/drivers/block/paride/pd.c
+++ b/drivers/block/paride/pd.c
@@ -943,7 +943,7 @@ static int pd_probe_drive(struct pd_unit *disk, int autoprobe, int port,
goto cleanup_disk;
return 0;
cleanup_disk:
- blk_cleanup_disk(disk->gd);
+ put_disk(disk->gd);
put_disk:
put_disk(p);
disk->gd = NULL;
@@ -1018,7 +1018,7 @@ static void __exit pd_exit(void)
if (p) {
disk->gd = NULL;
del_gendisk(p);
- blk_cleanup_disk(p);
+ put_disk(p);
blk_mq_free_tag_set(&disk->tag_set);
pi_release(disk->pi);
}
diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c
index 292e9a4ce1b9e..eec1b9fde2450 100644
--- a/drivers/block/paride/pf.c
+++ b/drivers/block/paride/pf.c
@@ -975,7 +975,7 @@ static int __init pf_init_unit(struct pf_unit *pf, bool autoprobe, int port,
out_pi_release:
pi_release(pf->pi);
out_free_disk:
- blk_cleanup_disk(pf->disk);
+ put_disk(pf->disk);
out_free_tag_set:
blk_mq_free_tag_set(&pf->tag_set);
return ret;
@@ -1044,7 +1044,7 @@ static void __exit pf_exit(void)
if (!pf->present)
continue;
del_gendisk(pf->disk);
- blk_cleanup_disk(pf->disk);
+ put_disk(pf->disk);
blk_mq_free_tag_set(&pf->tag_set);
pi_release(pf->pi);
}
diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c
index 7890933753443..653d242314830 100644
--- a/drivers/block/pktcdvd.c
+++ b/drivers/block/pktcdvd.c
@@ -2733,7 +2733,7 @@ static int pkt_setup_dev(dev_t dev, dev_t* pkt_dev)
return 0;
out_mem2:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_mem:
mempool_exit(&pd->rb_pool);
kfree(pd);
@@ -2783,7 +2783,7 @@ static int pkt_remove_dev(dev_t pkt_dev)
pkt_dbg(1, pd, "writer unmapped\n");
del_gendisk(pd->disk);
- blk_cleanup_disk(pd->disk);
+ put_disk(pd->disk);
mempool_exit(&pd->rb_pool);
kfree(pd);
diff --git a/drivers/block/ps3disk.c b/drivers/block/ps3disk.c
index 3054adf774603..36d7b36c60c76 100644
--- a/drivers/block/ps3disk.c
+++ b/drivers/block/ps3disk.c
@@ -473,7 +473,7 @@ static int ps3disk_probe(struct ps3_system_bus_device *_dev)
return 0;
fail_cleanup_disk:
- blk_cleanup_disk(gendisk);
+ put_disk(gendisk);
fail_free_tag_set:
blk_mq_free_tag_set(&priv->tag_set);
fail_teardown:
@@ -500,7 +500,7 @@ static void ps3disk_remove(struct ps3_system_bus_device *_dev)
&ps3disk_mask);
mutex_unlock(&ps3disk_mask_mutex);
del_gendisk(priv->gendisk);
- blk_cleanup_disk(priv->gendisk);
+ put_disk(priv->gendisk);
blk_mq_free_tag_set(&priv->tag_set);
dev_notice(&dev->sbd.core, "Synchronizing disk cache\n");
ps3disk_sync_cache(dev);
diff --git a/drivers/block/ps3vram.c b/drivers/block/ps3vram.c
index 4f90819e245e9..d1e0fefec90ba 100644
--- a/drivers/block/ps3vram.c
+++ b/drivers/block/ps3vram.c
@@ -761,7 +761,7 @@ static int ps3vram_probe(struct ps3_system_bus_device *dev)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(gendisk);
+ put_disk(gendisk);
out_cache_cleanup:
remove_proc_entry(DEVICE_NAME, NULL);
ps3vram_cache_cleanup(dev);
@@ -792,7 +792,7 @@ static void ps3vram_remove(struct ps3_system_bus_device *dev)
struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev);
del_gendisk(priv->gendisk);
- blk_cleanup_disk(priv->gendisk);
+ put_disk(priv->gendisk);
remove_proc_entry(DEVICE_NAME, NULL);
ps3vram_cache_cleanup(dev);
iounmap(priv->reports);
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c
index ef9bc62e9afd7..0d8ec2fe57400 100644
--- a/drivers/block/rbd.c
+++ b/drivers/block/rbd.c
@@ -4729,7 +4729,7 @@ static blk_status_t rbd_queue_rq(struct blk_mq_hw_ctx *hctx,
static void rbd_free_disk(struct rbd_device *rbd_dev)
{
- blk_cleanup_disk(rbd_dev->disk);
+ put_disk(rbd_dev->disk);
blk_mq_free_tag_set(&rbd_dev->tag_set);
rbd_dev->disk = NULL;
}
diff --git a/drivers/block/rnbd/rnbd-clt.c b/drivers/block/rnbd/rnbd-clt.c
index a4470374f54fc..b8d9e2824d9c7 100644
--- a/drivers/block/rnbd/rnbd-clt.c
+++ b/drivers/block/rnbd/rnbd-clt.c
@@ -1408,7 +1408,7 @@ static int rnbd_clt_setup_gen_disk(struct rnbd_clt_dev *dev, int idx)
blk_queue_flag_set(QUEUE_FLAG_NONROT, dev->queue);
err = add_disk(dev->gd);
if (err)
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
return err;
}
@@ -1630,7 +1630,7 @@ struct rnbd_clt_dev *rnbd_clt_map_device(const char *sessname,
static void destroy_gen_disk(struct rnbd_clt_dev *dev)
{
del_gendisk(dev->gd);
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
}
static void destroy_sysfs(struct rnbd_clt_dev *dev,
diff --git a/drivers/block/sunvdc.c b/drivers/block/sunvdc.c
index dd0a1a6fed296..fb855da971ee7 100644
--- a/drivers/block/sunvdc.c
+++ b/drivers/block/sunvdc.c
@@ -886,7 +886,7 @@ static int probe_disk(struct vdc_port *port)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(g);
+ put_disk(g);
out_free_tag:
blk_mq_free_tag_set(&port->tag_set);
return err;
@@ -1070,7 +1070,7 @@ static void vdc_port_remove(struct vio_dev *vdev)
del_timer_sync(&port->vio.timer);
del_gendisk(port->disk);
- blk_cleanup_disk(port->disk);
+ put_disk(port->disk);
blk_mq_free_tag_set(&port->tag_set);
vdc_free_tx_ring(port);
diff --git a/drivers/block/swim.c b/drivers/block/swim.c
index fef65a18d56fa..42b4b68286909 100644
--- a/drivers/block/swim.c
+++ b/drivers/block/swim.c
@@ -783,7 +783,7 @@ static void swim_cleanup_floppy_disk(struct floppy_state *fs)
if (fs->registered)
del_gendisk(fs->disk);
- blk_cleanup_disk(disk);
+ put_disk(disk);
blk_mq_free_tag_set(&fs->tag_set);
}
diff --git a/drivers/block/swim3.c b/drivers/block/swim3.c
index 6c39f2c9f806d..da811a7da03f3 100644
--- a/drivers/block/swim3.c
+++ b/drivers/block/swim3.c
@@ -1238,7 +1238,7 @@ static int swim3_attach(struct macio_dev *mdev,
return 0;
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_free_tag_set:
blk_mq_free_tag_set(&fs->tag_set);
out_unregister:
diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c
index 75057dbbcfbea..0e1a484cab0b6 100644
--- a/drivers/block/sx8.c
+++ b/drivers/block/sx8.c
@@ -1377,7 +1377,7 @@ static void carm_free_disk(struct carm_host *host, unsigned int port_no)
if (host->state > HST_DEV_ACTIVATE)
del_gendisk(disk);
- blk_cleanup_disk(disk);
+ put_disk(disk);
}
static int carm_init_shm(struct carm_host *host)
diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c
index cff1b6f6b0548..d7d72e8f6e551 100644
--- a/drivers/block/virtio_blk.c
+++ b/drivers/block/virtio_blk.c
@@ -1089,7 +1089,7 @@ static int virtblk_probe(struct virtio_device *vdev)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(vblk->disk);
+ put_disk(vblk->disk);
out_free_tags:
blk_mq_free_tag_set(&vblk->tag_set);
out_free_vq:
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c
index a88ce4426400e..ee909d7a50fc9 100644
--- a/drivers/block/xen-blkfront.c
+++ b/drivers/block/xen-blkfront.c
@@ -2382,7 +2382,7 @@ static void blkfront_connect(struct blkfront_info *info)
err = device_add_disk(&info->xbdev->dev, info->gd, NULL);
if (err) {
- blk_cleanup_disk(info->gd);
+ put_disk(info->gd);
blk_mq_free_tag_set(&info->tag_set);
info->rq = NULL;
goto fail;
@@ -2465,7 +2465,7 @@ static int blkfront_remove(struct xenbus_device *xbdev)
blkif_free(info, 0);
xlbd_release_minors(info->gd->first_minor, info->gd->minors);
- blk_cleanup_disk(info->gd);
+ put_disk(info->gd);
blk_mq_free_tag_set(&info->tag_set);
kfree(info);
diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c
index 18ad43d9933ec..c1e85f356e4df 100644
--- a/drivers/block/z2ram.c
+++ b/drivers/block/z2ram.c
@@ -337,7 +337,7 @@ static int z2ram_register_disk(int minor)
z2ram_gendisk[minor] = disk;
err = add_disk(disk);
if (err)
- blk_cleanup_disk(disk);
+ put_disk(disk);
return err;
}
diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
index b8549c61ff2ce..e5233c911e436 100644
--- a/drivers/block/zram/zram_drv.c
+++ b/drivers/block/zram/zram_drv.c
@@ -1957,7 +1957,7 @@ static int zram_add(void)
return device_id;
out_cleanup_disk:
- blk_cleanup_disk(zram->disk);
+ put_disk(zram->disk);
out_free_idr:
idr_remove(&zram_index_idr, device_id);
out_free_dev:
@@ -2008,7 +2008,7 @@ static int zram_remove(struct zram *zram)
*/
zram_reset_device(zram);
- blk_cleanup_disk(zram->disk);
+ put_disk(zram->disk);
kfree(zram);
return 0;
}
diff --git a/drivers/cdrom/gdrom.c b/drivers/cdrom/gdrom.c
index f4cc90ea6198e..ceded5772aac6 100644
--- a/drivers/cdrom/gdrom.c
+++ b/drivers/cdrom/gdrom.c
@@ -817,7 +817,7 @@ static int probe_gdrom(struct platform_device *devptr)
free_irq(HW_EVENT_GDROM_DMA, &gd);
free_irq(HW_EVENT_GDROM_CMD, &gd);
probe_fail_cleanup_disk:
- blk_cleanup_disk(gd.disk);
+ put_disk(gd.disk);
probe_fail_free_tag_set:
blk_mq_free_tag_set(&gd.tag_set);
probe_fail_free_cd_info:
diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index 3563d15dbaf27..9dd752d272f60 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -884,7 +884,7 @@ static void bcache_device_free(struct bcache_device *d)
if (disk) {
ida_simple_remove(&bcache_device_idx,
first_minor_to_idx(disk->first_minor));
- blk_cleanup_disk(disk);
+ put_disk(disk);
}
bioset_exit(&d->bio_split);
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index d8f16183bf27c..a46474c4900a2 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -1884,7 +1884,7 @@ static void cleanup_mapped_device(struct mapped_device *md)
del_gendisk(md->disk);
}
dm_queue_destroy_crypto_profile(md->queue);
- blk_cleanup_disk(md->disk);
+ put_disk(md->disk);
}
if (md->pending_io) {
diff --git a/drivers/md/md.c b/drivers/md/md.c
index c7ecb0bffda0d..076255ec9ba18 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -5579,7 +5579,7 @@ static void md_free(struct kobject *ko)
if (mddev->gendisk) {
del_gendisk(mddev->gendisk);
- blk_cleanup_disk(mddev->gendisk);
+ put_disk(mddev->gendisk);
}
percpu_ref_exit(&mddev->writes_pending);
@@ -5718,7 +5718,7 @@ static int md_alloc(dev_t dev, char *name)
out_del_gendisk:
del_gendisk(disk);
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_unlock_disks_mutex:
mutex_unlock(&disks_mutex);
mddev_put(mddev);
diff --git a/drivers/memstick/core/ms_block.c b/drivers/memstick/core/ms_block.c
index ba7e7249a3db4..ed9a683b3ca86 100644
--- a/drivers/memstick/core/ms_block.c
+++ b/drivers/memstick/core/ms_block.c
@@ -2129,7 +2129,7 @@ static int msb_init_disk(struct memstick_dev *card)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(msb->disk);
+ put_disk(msb->disk);
out_free_tag_set:
blk_mq_free_tag_set(&msb->tag_set);
out_release_id:
diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c
index 72e91c06c618b..61cf75d4a01ea 100644
--- a/drivers/memstick/core/mspro_block.c
+++ b/drivers/memstick/core/mspro_block.c
@@ -1209,7 +1209,7 @@ static int mspro_block_init_disk(struct memstick_dev *card)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(msb->disk);
+ put_disk(msb->disk);
out_free_tag_set:
blk_mq_free_tag_set(&msb->tag_set);
out_release_id:
diff --git a/drivers/mtd/mtd_blkdevs.c b/drivers/mtd/mtd_blkdevs.c
index f731721114655..60b222799871e 100644
--- a/drivers/mtd/mtd_blkdevs.c
+++ b/drivers/mtd/mtd_blkdevs.c
@@ -29,7 +29,7 @@ static void blktrans_dev_release(struct kref *kref)
struct mtd_blktrans_dev *dev =
container_of(kref, struct mtd_blktrans_dev, ref);
- blk_cleanup_disk(dev->disk);
+ put_disk(dev->disk);
blk_mq_free_tag_set(dev->tag_set);
kfree(dev->tag_set);
list_del(&dev->list);
@@ -398,7 +398,7 @@ int add_mtd_blktrans_dev(struct mtd_blktrans_dev *new)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(new->disk);
+ put_disk(new->disk);
out_free_tag_set:
blk_mq_free_tag_set(new->tag_set);
out_kfree_tag_set:
diff --git a/drivers/mtd/ubi/block.c b/drivers/mtd/ubi/block.c
index a78fdf3b30f7e..4cf67a2a0d04b 100644
--- a/drivers/mtd/ubi/block.c
+++ b/drivers/mtd/ubi/block.c
@@ -467,7 +467,7 @@ int ubiblock_create(struct ubi_volume_info *vi)
out_remove_minor:
idr_remove(&ubiblock_minor_idr, gd->first_minor);
out_cleanup_disk:
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
out_free_tags:
blk_mq_free_tag_set(&dev->tag_set);
out_free_dev:
@@ -486,7 +486,7 @@ static void ubiblock_cleanup(struct ubiblock *dev)
destroy_workqueue(dev->wq);
/* Finally destroy the blk queue */
dev_info(disk_to_dev(dev->gd), "released");
- blk_cleanup_disk(dev->gd);
+ put_disk(dev->gd);
blk_mq_free_tag_set(&dev->tag_set);
idr_remove(&ubiblock_minor_idr, dev->gd->first_minor);
}
diff --git a/drivers/nvdimm/btt.c b/drivers/nvdimm/btt.c
index 9613e54c7a675..5e622c0d4b66a 100644
--- a/drivers/nvdimm/btt.c
+++ b/drivers/nvdimm/btt.c
@@ -1548,14 +1548,14 @@ static int btt_blk_init(struct btt *btt)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(btt->btt_disk);
+ put_disk(btt->btt_disk);
return rc;
}
static void btt_blk_cleanup(struct btt *btt)
{
del_gendisk(btt->btt_disk);
- blk_cleanup_disk(btt->btt_disk);
+ put_disk(btt->btt_disk);
}
/**
diff --git a/drivers/nvdimm/pmem.c b/drivers/nvdimm/pmem.c
index 629d10fcf53b2..a72b81fa32426 100644
--- a/drivers/nvdimm/pmem.c
+++ b/drivers/nvdimm/pmem.c
@@ -450,7 +450,7 @@ static void pmem_release_disk(void *__pmem)
put_dax(pmem->dax_dev);
del_gendisk(pmem->disk);
- blk_cleanup_disk(pmem->disk);
+ put_disk(pmem->disk);
}
static int pmem_attach_disk(struct device *dev,
@@ -596,7 +596,7 @@ static int pmem_attach_disk(struct device *dev,
kill_dax(pmem->dax_dev);
put_dax(pmem->dax_dev);
out:
- blk_cleanup_disk(pmem->disk);
+ put_disk(pmem->disk);
return rc;
}
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 697af049b49d4..e14413cfe95c7 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -4047,7 +4047,7 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, unsigned nsid,
mutex_unlock(&ctrl->subsys->lock);
nvme_put_ns_head(ns->head);
out_cleanup_disk:
- blk_cleanup_disk(disk);
+ put_disk(disk);
out_free_ns:
kfree(ns);
out_free_id:
diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index d3e2440d8abb0..ccf9a6da8f6e1 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -853,7 +853,7 @@ void nvme_mpath_remove_disk(struct nvme_ns_head *head)
/* make sure all pending bios are cleaned up */
kblockd_schedule_work(&head->requeue_work);
flush_work(&head->requeue_work);
- blk_cleanup_disk(head->disk);
+ put_disk(head->disk);
}
void nvme_mpath_init_ctrl(struct nvme_ctrl *ctrl)
diff --git a/drivers/s390/block/dcssblk.c b/drivers/s390/block/dcssblk.c
index 8d0d0eaa30594..4d8d1759775ae 100644
--- a/drivers/s390/block/dcssblk.c
+++ b/drivers/s390/block/dcssblk.c
@@ -414,7 +414,7 @@ dcssblk_shared_store(struct device *dev, struct device_attribute *attr, const ch
kill_dax(dev_info->dax_dev);
put_dax(dev_info->dax_dev);
del_gendisk(dev_info->gd);
- blk_cleanup_disk(dev_info->gd);
+ put_disk(dev_info->gd);
up_write(&dcssblk_devices_sem);
if (device_remove_file_self(dev, attr)) {
@@ -712,7 +712,7 @@ dcssblk_add_store(struct device *dev, struct device_attribute *attr, const char
put_dax(dev_info->dax_dev);
put_dev:
list_del(&dev_info->lh);
- blk_cleanup_disk(dev_info->gd);
+ put_disk(dev_info->gd);
list_for_each_entry(seg_info, &dev_info->seg_list, lh) {
segment_unload(seg_info->segment_name);
}
@@ -722,7 +722,7 @@ dcssblk_add_store(struct device *dev, struct device_attribute *attr, const char
dev_list_del:
list_del(&dev_info->lh);
release_gd:
- blk_cleanup_disk(dev_info->gd);
+ put_disk(dev_info->gd);
up_write(&dcssblk_devices_sem);
seg_list_del:
if (dev_info == NULL)
@@ -790,7 +790,7 @@ dcssblk_remove_store(struct device *dev, struct device_attribute *attr, const ch
kill_dax(dev_info->dax_dev);
put_dax(dev_info->dax_dev);
del_gendisk(dev_info->gd);
- blk_cleanup_disk(dev_info->gd);
+ put_disk(dev_info->gd);
/* unload all related segments */
list_for_each_entry(entry, &dev_info->seg_list, lh)
diff --git a/drivers/s390/block/scm_blk.c b/drivers/s390/block/scm_blk.c
index 2a9c0ddcade59..0c1df1d5f1aca 100644
--- a/drivers/s390/block/scm_blk.c
+++ b/drivers/s390/block/scm_blk.c
@@ -501,7 +501,7 @@ int scm_blk_dev_setup(struct scm_blk_dev *bdev, struct scm_device *scmdev)
return 0;
out_cleanup_disk:
- blk_cleanup_disk(bdev->gendisk);
+ put_disk(bdev->gendisk);
out_tag:
blk_mq_free_tag_set(&bdev->tag_set);
out:
@@ -512,7 +512,7 @@ int scm_blk_dev_setup(struct scm_blk_dev *bdev, struct scm_device *scmdev)
void scm_blk_dev_cleanup(struct scm_blk_dev *bdev)
{
del_gendisk(bdev->gendisk);
- blk_cleanup_disk(bdev->gendisk);
+ put_disk(bdev->gendisk);
blk_mq_free_tag_set(&bdev->tag_set);
}
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index b9bf092531bc7..cfe36c87ac4ca 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -830,7 +830,6 @@ struct gendisk *__blk_alloc_disk(int node, struct lock_class_key *lkclass);
\
__blk_alloc_disk(node_id, &__key); \
})
-void blk_cleanup_disk(struct gendisk *disk);
int __register_blkdev(unsigned int major, const char *name,
void (*probe)(dev_t devt));
--
2.30.2
^ permalink raw reply related [flat|nested] 25+ messages in thread
* Re: [PATCH 3/6] block: remove QUEUE_FLAG_DEAD
2022-06-19 6:05 ` [PATCH 3/6] block: remove QUEUE_FLAG_DEAD Christoph Hellwig
@ 2022-06-19 14:07 ` Bart Van Assche
2022-06-20 6:20 ` Christoph Hellwig
2022-06-20 8:39 ` Hannes Reinecke
1 sibling, 1 reply; 25+ messages in thread
From: Bart Van Assche @ 2022-06-19 14:07 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/18/22 23:05, Christoph Hellwig wrote:
> @@ -151,11 +150,10 @@ static ssize_t queue_state_write(void *data, const char __user *buf,
> char opbuf[16] = { }, *op;
>
> /*
> - * The "state" attribute is removed after blk_cleanup_queue() has called
> - * blk_mq_free_queue(). Return if QUEUE_FLAG_DEAD has been set to avoid
> - * triggering a use-after-free.
> + * The "state" attribute is removed when the queue is removed. Don't
> + * allow setting the state on a dying queue to avoid a use-after-free.
> */
> - if (blk_queue_dead(q))
> + if (blk_queue_dying(q))
> return -ENOENT;
I'm missing an explanation of why this patch forbids triggering a queue
run in the dying state. "dying" means that allocation of new requests
will fail. Unless if something fundamentally has changed in the block
layer it should still be safe to trigger a queue run in the "dying" state.
Thanks,
Bart.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
` (5 preceding siblings ...)
2022-06-19 6:05 ` [PATCH 6/6] block: remove blk_cleanup_disk Christoph Hellwig
@ 2022-06-19 22:21 ` Jens Axboe
2022-06-20 6:09 ` Christoph Hellwig
2022-06-28 12:33 ` Jens Axboe
7 siblings, 1 reply; 25+ messages in thread
From: Jens Axboe @ 2022-06-19 22:21 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Ming Lei, linux-block
On 6/19/22 12:05 AM, Christoph Hellwig wrote:
> Note that while intended or 5.20, this series is generated against the
> block-5.19 branch as that contains fixes in this area that haven't
> made it to the for-5.10/block branch yet.
Side note - I rebased on -rc3 anyway because of the series that went
into -rc2, so we should be fine there.
--
Jens Axboe
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-19 22:21 ` fully tear down the queue in del_gendisk Jens Axboe
@ 2022-06-20 6:09 ` Christoph Hellwig
2022-06-20 11:16 ` Jens Axboe
0 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-20 6:09 UTC (permalink / raw)
To: Jens Axboe; +Cc: Christoph Hellwig, Ming Lei, linux-block
On Sun, Jun 19, 2022 at 04:21:39PM -0600, Jens Axboe wrote:
> On 6/19/22 12:05 AM, Christoph Hellwig wrote:
> > Note that while intended or 5.20, this series is generated against the
> > block-5.19 branch as that contains fixes in this area that haven't
> > made it to the for-5.10/block branch yet.
>
> Side note - I rebased on -rc3 anyway because of the series that went
> into -rc2, so we should be fine there.
It depends on elevator/debugfs teardown series that has not made it
into -rc3.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 3/6] block: remove QUEUE_FLAG_DEAD
2022-06-19 14:07 ` Bart Van Assche
@ 2022-06-20 6:20 ` Christoph Hellwig
0 siblings, 0 replies; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-20 6:20 UTC (permalink / raw)
To: Bart Van Assche; +Cc: Christoph Hellwig, Jens Axboe, Ming Lei, linux-block
On Sun, Jun 19, 2022 at 07:07:20AM -0700, Bart Van Assche wrote:
>> - if (blk_queue_dead(q))
>> + if (blk_queue_dying(q))
>> return -ENOENT;
>
> I'm missing an explanation of why this patch forbids triggering a queue run
> in the dying state. "dying" means that allocation of new requests will
> fail. Unless if something fundamentally has changed in the block layer it
> should still be safe to trigger a queue run in the "dying" state.
It is safe, but not worth having another queue state for given that we
can otherwise remove the QUEUE_FLAG_DEAD bit.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 1/6] mtip32xx: remove the device_status debugfs file
2022-06-19 6:05 ` [PATCH 1/6] mtip32xx: remove the device_status debugfs file Christoph Hellwig
@ 2022-06-20 8:37 ` Hannes Reinecke
0 siblings, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:37 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> This file is a huge mess that iterates over all devices and is in the
> way of fixing the device removal in this driver, so remove it.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> drivers/block/mtip32xx/mtip32xx.c | 141 +-----------------------------
> drivers/block/mtip32xx/mtip32xx.h | 4 -
> 2 files changed, 1 insertion(+), 144 deletions(-)
>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 2/6] mtip32xx: fix device removal
2022-06-19 6:05 ` [PATCH 2/6] mtip32xx: fix device removal Christoph Hellwig
@ 2022-06-20 8:38 ` Hannes Reinecke
0 siblings, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:38 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> Use the proper helper to mark a surpise removal, remove the gendisk as
> soon as possible when removing the device and implement the ->free_disk
> callback to ensure the private data is alive as long as the gendisk has
> references.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> drivers/block/mtip32xx/mtip32xx.c | 157 +++++++++---------------------
> drivers/block/mtip32xx/mtip32xx.h | 1 -
> 2 files changed, 44 insertions(+), 114 deletions(-)
>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 3/6] block: remove QUEUE_FLAG_DEAD
2022-06-19 6:05 ` [PATCH 3/6] block: remove QUEUE_FLAG_DEAD Christoph Hellwig
2022-06-19 14:07 ` Bart Van Assche
@ 2022-06-20 8:39 ` Hannes Reinecke
1 sibling, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:39 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> Disallow setting the blk-mq state on any queue that is already dying as
> setting the state even then is a bad idea, and remove the now unused
> QUEUE_FLAG_DEAD flag.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> block/blk-core.c | 3 ---
> block/blk-mq-debugfs.c | 8 +++-----
> include/linux/blkdev.h | 2 --
> 3 files changed, 3 insertions(+), 10 deletions(-)
>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue
2022-06-19 6:05 ` [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue Christoph Hellwig
@ 2022-06-20 8:40 ` Hannes Reinecke
0 siblings, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:40 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> These flags only apply to file system I/O, and all file system I/O is
> already drained by del_gendisk and thus can't be in progress when
> blk_cleanup_queue is called.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> block/blk-core.c | 3 ---
> 1 file changed, 3 deletions(-)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index 088332984cd1b..2f418606e3bd3 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -304,9 +304,6 @@ void blk_cleanup_queue(struct request_queue *q)
> blk_queue_flag_set(QUEUE_FLAG_DYING, q);
> blk_queue_start_drain(q);
>
> - blk_queue_flag_set(QUEUE_FLAG_NOMERGES, q);
> - blk_queue_flag_set(QUEUE_FLAG_NOXMERGES, q);
> -
> /*
> * Drain all requests queued before DYING marking. Set DEAD flag to
> * prevent that blk_mq_run_hw_queues() accesses the hardware queues
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 5/6] block: simplify disk shutdown
2022-06-19 6:05 ` [PATCH 5/6] block: simplify disk shutdown Christoph Hellwig
@ 2022-06-20 8:47 ` Hannes Reinecke
0 siblings, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:47 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> Set the queue dying flag and call blk_mq_exit_queue from del_gendisk for
> all disks that do not have separately allocated queues, and thus remove
> the need to call blk_cleanup_queue for them.
>
> Rename blk_cleanup_disk to blk_mq_destroy_queue to make it clear that
> this function is intended only for separately allocated blk-mq queues.
>
> This saves an extra queue freeze for devices without a separately
> allocated queue.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> block/blk-core.c | 37 -------------------------
> block/blk-mq.c | 43 +++++++++++++++++++++++++++--
> block/blk-sysfs.c | 5 ----
> block/blk.h | 3 ++
> block/bsg-lib.c | 4 +--
> block/genhd.c | 23 ++++++++-------
> drivers/block/ataflop.c | 1 -
> drivers/block/loop.c | 1 -
> drivers/block/mtip32xx/mtip32xx.c | 2 --
> drivers/block/rnbd/rnbd-clt.c | 2 +-
> drivers/block/sx8.c | 4 +--
> drivers/block/virtio_blk.c | 1 -
> drivers/block/z2ram.c | 1 -
> drivers/cdrom/gdrom.c | 1 -
> drivers/memstick/core/ms_block.c | 1 -
> drivers/memstick/core/mspro_block.c | 1 -
> drivers/mmc/core/block.c | 1 -
> drivers/mmc/core/queue.c | 1 -
> drivers/nvme/host/apple.c | 2 +-
> drivers/nvme/host/core.c | 1 -
> drivers/nvme/host/fc.c | 12 ++++----
> drivers/nvme/host/pci.c | 2 +-
> drivers/nvme/host/rdma.c | 12 ++++----
> drivers/nvme/host/tcp.c | 12 ++++----
> drivers/nvme/target/loop.c | 12 ++++----
> drivers/s390/block/dasd.c | 2 +-
> drivers/s390/block/dasd_genhd.c | 4 +--
> drivers/scsi/scsi_lib.c | 6 ++--
> drivers/scsi/scsi_sysfs.c | 2 +-
> drivers/scsi/sd.c | 4 +--
> drivers/scsi/sr.c | 4 +--
> drivers/ufs/core/ufshcd.c | 4 +--
> include/linux/blk-mq.h | 3 ++
> include/linux/blkdev.h | 4 +--
> 34 files changed, 105 insertions(+), 113 deletions(-)
>
(This will come back to haunt me once I have to do a backport ...)
(But anyway.)
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 6/6] block: remove blk_cleanup_disk
2022-06-19 6:05 ` [PATCH 6/6] block: remove blk_cleanup_disk Christoph Hellwig
@ 2022-06-20 8:49 ` Hannes Reinecke
2022-06-20 8:56 ` Christoph Hellwig
0 siblings, 1 reply; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 8:49 UTC (permalink / raw)
To: Christoph Hellwig, Jens Axboe; +Cc: Ming Lei, linux-block
On 6/19/22 08:05, Christoph Hellwig wrote:
> blk_cleanup_disk is nothing but a trivial wrapper for put_disk now,
> so remove it.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
> arch/m68k/emu/nfblock.c | 4 ++--
> arch/um/drivers/ubd_kern.c | 4 ++--
> arch/xtensa/platforms/iss/simdisk.c | 4 ++--
> block/genhd.c | 15 ---------------
> drivers/block/amiflop.c | 2 +-
> drivers/block/aoe/aoeblk.c | 2 +-
> drivers/block/aoe/aoedev.c | 2 +-
> drivers/block/ataflop.c | 4 ++--
> drivers/block/brd.c | 4 ++--
> drivers/block/drbd/drbd_main.c | 4 ++--
> drivers/block/floppy.c | 6 +++---
> drivers/block/loop.c | 2 +-
> drivers/block/mtip32xx/mtip32xx.c | 2 +-
> drivers/block/n64cart.c | 2 +-
> drivers/block/nbd.c | 4 ++--
> drivers/block/null_blk/main.c | 4 ++--
> drivers/block/paride/pcd.c | 4 ++--
> drivers/block/paride/pd.c | 4 ++--
> drivers/block/paride/pf.c | 4 ++--
> drivers/block/pktcdvd.c | 4 ++--
> drivers/block/ps3disk.c | 4 ++--
> drivers/block/ps3vram.c | 4 ++--
> drivers/block/rbd.c | 2 +-
> drivers/block/rnbd/rnbd-clt.c | 4 ++--
> drivers/block/sunvdc.c | 4 ++--
> drivers/block/swim.c | 2 +-
> drivers/block/swim3.c | 2 +-
> drivers/block/sx8.c | 2 +-
> drivers/block/virtio_blk.c | 2 +-
> drivers/block/xen-blkfront.c | 4 ++--
> drivers/block/z2ram.c | 2 +-
> drivers/block/zram/zram_drv.c | 4 ++--
> drivers/cdrom/gdrom.c | 2 +-
> drivers/md/bcache/super.c | 2 +-
> drivers/md/dm.c | 2 +-
> drivers/md/md.c | 4 ++--
> drivers/memstick/core/ms_block.c | 2 +-
> drivers/memstick/core/mspro_block.c | 2 +-
> drivers/mtd/mtd_blkdevs.c | 4 ++--
> drivers/mtd/ubi/block.c | 4 ++--
> drivers/nvdimm/btt.c | 4 ++--
> drivers/nvdimm/pmem.c | 4 ++--
> drivers/nvme/host/core.c | 2 +-
> drivers/nvme/host/multipath.c | 2 +-
> drivers/s390/block/dcssblk.c | 8 ++++----
> drivers/s390/block/scm_blk.c | 4 ++--
> include/linux/blkdev.h | 1 -
> 47 files changed, 74 insertions(+), 90 deletions(-)
>
I wish we could have blktests for tearing down device-drivers; doing a
regression test here will be really hard.
Reviewed-by: Hannes Reinecke <hare@suse.de>
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 6/6] block: remove blk_cleanup_disk
2022-06-20 8:49 ` Hannes Reinecke
@ 2022-06-20 8:56 ` Christoph Hellwig
2022-06-20 9:00 ` Hannes Reinecke
2022-06-28 14:42 ` Sagi Grimberg
0 siblings, 2 replies; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-20 8:56 UTC (permalink / raw)
To: Hannes Reinecke; +Cc: Christoph Hellwig, Jens Axboe, Ming Lei, linux-block
On Mon, Jun 20, 2022 at 10:49:06AM +0200, Hannes Reinecke wrote:
> I wish we could have blktests for tearing down device-drivers; doing a
> regression test here will be really hard.
The problem with a remove is that while we have a generic device remove
attribute, it:
a) isn't always in the same place relatively to the disk
b) once removed we have no generic way to add the device back for
further testing
nvme/032 has basic remove testing for nvme, and I think I can also
wire up my scsi bind/unbind testing for blktests using scsi_debug without
too much effort. But that still isn't exactly a generic test.
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 6/6] block: remove blk_cleanup_disk
2022-06-20 8:56 ` Christoph Hellwig
@ 2022-06-20 9:00 ` Hannes Reinecke
2022-06-28 14:42 ` Sagi Grimberg
1 sibling, 0 replies; 25+ messages in thread
From: Hannes Reinecke @ 2022-06-20 9:00 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Jens Axboe, Ming Lei, linux-block
On 6/20/22 10:56, Christoph Hellwig wrote:
> On Mon, Jun 20, 2022 at 10:49:06AM +0200, Hannes Reinecke wrote:
>> I wish we could have blktests for tearing down device-drivers; doing a
>> regression test here will be really hard.
>
> The problem with a remove is that while we have a generic device remove
> attribute, it:
>
> a) isn't always in the same place relatively to the disk
> b) once removed we have no generic way to add the device back for
> further testing
>
> nvme/032 has basic remove testing for nvme, and I think I can also
> wire up my scsi bind/unbind testing for blktests using scsi_debug without
> too much effort. But that still isn't exactly a generic test.
Precisely my sentiment. This patchset is touch a _lot_ of drivers, and
modifying the removal path of each.
Looking obvious from initial glance, but as usual the devil's in the
detail. Or, to quote my QA maxim:
'The only way to be ensure it works is to ensure it works.'
But anyway, probably just wishful thinking.
Cheers,
Hannes
--
Dr. Hannes Reinecke Kernel Storage Architect
hare@suse.de +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-20 6:09 ` Christoph Hellwig
@ 2022-06-20 11:16 ` Jens Axboe
2022-06-28 5:11 ` Christoph Hellwig
0 siblings, 1 reply; 25+ messages in thread
From: Jens Axboe @ 2022-06-20 11:16 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Ming Lei, linux-block
On 6/20/22 12:09 AM, Christoph Hellwig wrote:
> On Sun, Jun 19, 2022 at 04:21:39PM -0600, Jens Axboe wrote:
>> On 6/19/22 12:05 AM, Christoph Hellwig wrote:
>>> Note that while intended or 5.20, this series is generated against the
>>> block-5.19 branch as that contains fixes in this area that haven't
>>> made it to the for-5.10/block branch yet.
>>
>> Side note - I rebased on -rc3 anyway because of the series that went
>> into -rc2, so we should be fine there.
>
> It depends on elevator/debugfs teardown series that has not made it
> into -rc3.
Guess we'll rebase for -rc4 again...
--
Jens Axboe
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-20 11:16 ` Jens Axboe
@ 2022-06-28 5:11 ` Christoph Hellwig
2022-06-28 12:29 ` Jens Axboe
0 siblings, 1 reply; 25+ messages in thread
From: Christoph Hellwig @ 2022-06-28 5:11 UTC (permalink / raw)
To: Jens Axboe; +Cc: Christoph Hellwig, Ming Lei, linux-block
On Mon, Jun 20, 2022 at 05:16:03AM -0600, Jens Axboe wrote:
> On 6/20/22 12:09 AM, Christoph Hellwig wrote:
> > On Sun, Jun 19, 2022 at 04:21:39PM -0600, Jens Axboe wrote:
> >> On 6/19/22 12:05 AM, Christoph Hellwig wrote:
> >>> Note that while intended or 5.20, this series is generated against the
> >>> block-5.19 branch as that contains fixes in this area that haven't
> >>> made it to the for-5.10/block branch yet.
> >>
> >> Side note - I rebased on -rc3 anyway because of the series that went
> >> into -rc2, so we should be fine there.
> >
> > It depends on elevator/debugfs teardown series that has not made it
> > into -rc3.
>
> Guess we'll rebase for -rc4 again...
As you'd done that, can you consider this series now?
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-28 5:11 ` Christoph Hellwig
@ 2022-06-28 12:29 ` Jens Axboe
2022-06-28 12:35 ` Jens Axboe
0 siblings, 1 reply; 25+ messages in thread
From: Jens Axboe @ 2022-06-28 12:29 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Ming Lei, linux-block
On 6/27/22 11:11 PM, Christoph Hellwig wrote:
> On Mon, Jun 20, 2022 at 05:16:03AM -0600, Jens Axboe wrote:
>> On 6/20/22 12:09 AM, Christoph Hellwig wrote:
>>> On Sun, Jun 19, 2022 at 04:21:39PM -0600, Jens Axboe wrote:
>>>> On 6/19/22 12:05 AM, Christoph Hellwig wrote:
>>>>> Note that while intended or 5.20, this series is generated against the
>>>>> block-5.19 branch as that contains fixes in this area that haven't
>>>>> made it to the for-5.10/block branch yet.
>>>>
>>>> Side note - I rebased on -rc3 anyway because of the series that went
>>>> into -rc2, so we should be fine there.
>>>
>>> It depends on elevator/debugfs teardown series that has not made it
>>> into -rc3.
>>
>> Guess we'll rebase for -rc4 again...
>
> As you'd done that, can you consider this series now?
It doesn't apply cleanly. I'll check if it's something minor...
--
Jens Axboe
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
` (6 preceding siblings ...)
2022-06-19 22:21 ` fully tear down the queue in del_gendisk Jens Axboe
@ 2022-06-28 12:33 ` Jens Axboe
7 siblings, 0 replies; 25+ messages in thread
From: Jens Axboe @ 2022-06-28 12:33 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: linux-block, ming.lei
On Sun, 19 Jun 2022 08:05:46 +0200, Christoph Hellwig wrote:
> this series simplifies teardown for most block drivers. Right now they
> also have to call blk_cleanup_queue after calling del_gendisk, making
> the teardown process rather confusing.
>
> Instead this series records if the request_queue is owned by the gendisk,
> which is always the case except for scsi and dasd or queues without a
> gendisk at all, and then does the entire teardown in del_gendisk.
>
> [...]
Applied, thanks!
[1/6] mtip32xx: remove the device_status debugfs file
commit: ec5263f422a3364442e0db2d9c2866d9154cbcc4
[2/6] mtip32xx: fix device removal
commit: e8b58ef09e84c15cf782b01cfc73cc5b1180d519
[3/6] block: remove QUEUE_FLAG_DEAD
commit: 1f90307e5f0d7bc9a336ead528f616a5df8e5944
[4/6] block: stop setting the nomerges flags in blk_cleanup_queue
commit: 0e3534022f26ae51f7cf28347a253230604b6f4e
[5/6] block: simplify disk shutdown
commit: 6f8191fdf41d3a53cc1d63fe2234e812c55a0092
[6/6] block: remove blk_cleanup_disk
commit: 8b9ab62662048a3274361c7e5f64037c2c133e2c
Best regards,
--
Jens Axboe
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: fully tear down the queue in del_gendisk
2022-06-28 12:29 ` Jens Axboe
@ 2022-06-28 12:35 ` Jens Axboe
0 siblings, 0 replies; 25+ messages in thread
From: Jens Axboe @ 2022-06-28 12:35 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Ming Lei, linux-block
On 6/28/22 6:29 AM, Jens Axboe wrote:
> On 6/27/22 11:11 PM, Christoph Hellwig wrote:
>> On Mon, Jun 20, 2022 at 05:16:03AM -0600, Jens Axboe wrote:
>>> On 6/20/22 12:09 AM, Christoph Hellwig wrote:
>>>> On Sun, Jun 19, 2022 at 04:21:39PM -0600, Jens Axboe wrote:
>>>>> On 6/19/22 12:05 AM, Christoph Hellwig wrote:
>>>>>> Note that while intended or 5.20, this series is generated against the
>>>>>> block-5.19 branch as that contains fixes in this area that haven't
>>>>>> made it to the for-5.10/block branch yet.
>>>>>
>>>>> Side note - I rebased on -rc3 anyway because of the series that went
>>>>> into -rc2, so we should be fine there.
>>>>
>>>> It depends on elevator/debugfs teardown series that has not made it
>>>> into -rc3.
>>>
>>> Guess we'll rebase for -rc4 again...
>>
>> As you'd done that, can you consider this series now?
>
> It doesn't apply cleanly. I'll check if it's something minor...
Minor thing in xen-blkfront for the last patch, hand edited the patch:
https://git.kernel.dk/cgit/linux-block/commit/?h=for-5.20/block&id=8b9ab62662048a3274361c7e5f64037c2c133e2c
--
Jens Axboe
^ permalink raw reply [flat|nested] 25+ messages in thread
* Re: [PATCH 6/6] block: remove blk_cleanup_disk
2022-06-20 8:56 ` Christoph Hellwig
2022-06-20 9:00 ` Hannes Reinecke
@ 2022-06-28 14:42 ` Sagi Grimberg
1 sibling, 0 replies; 25+ messages in thread
From: Sagi Grimberg @ 2022-06-28 14:42 UTC (permalink / raw)
To: Christoph Hellwig, Hannes Reinecke; +Cc: Jens Axboe, Ming Lei, linux-block
>> I wish we could have blktests for tearing down device-drivers; doing a
>> regression test here will be really hard.
>
> The problem with a remove is that while we have a generic device remove
> attribute, it:
>
> a) isn't always in the same place relatively to the disk
> b) once removed we have no generic way to add the device back for
> further testing
>
> nvme/032 has basic remove testing for nvme, and I think I can also
> wire up my scsi bind/unbind testing for blktests using scsi_debug without
> too much effort. But that still isn't exactly a generic test.
It should be trivial to add a similar test for fabrics with toggling
nvmet $ns/enable config knob...
^ permalink raw reply [flat|nested] 25+ messages in thread
end of thread, other threads:[~2022-06-28 14:42 UTC | newest]
Thread overview: 25+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-06-19 6:05 fully tear down the queue in del_gendisk Christoph Hellwig
2022-06-19 6:05 ` [PATCH 1/6] mtip32xx: remove the device_status debugfs file Christoph Hellwig
2022-06-20 8:37 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 2/6] mtip32xx: fix device removal Christoph Hellwig
2022-06-20 8:38 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 3/6] block: remove QUEUE_FLAG_DEAD Christoph Hellwig
2022-06-19 14:07 ` Bart Van Assche
2022-06-20 6:20 ` Christoph Hellwig
2022-06-20 8:39 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 4/6] block: stop setting the nomerges flags in blk_cleanup_queue Christoph Hellwig
2022-06-20 8:40 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 5/6] block: simplify disk shutdown Christoph Hellwig
2022-06-20 8:47 ` Hannes Reinecke
2022-06-19 6:05 ` [PATCH 6/6] block: remove blk_cleanup_disk Christoph Hellwig
2022-06-20 8:49 ` Hannes Reinecke
2022-06-20 8:56 ` Christoph Hellwig
2022-06-20 9:00 ` Hannes Reinecke
2022-06-28 14:42 ` Sagi Grimberg
2022-06-19 22:21 ` fully tear down the queue in del_gendisk Jens Axboe
2022-06-20 6:09 ` Christoph Hellwig
2022-06-20 11:16 ` Jens Axboe
2022-06-28 5:11 ` Christoph Hellwig
2022-06-28 12:29 ` Jens Axboe
2022-06-28 12:35 ` Jens Axboe
2022-06-28 12:33 ` Jens Axboe
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.