null_blk: oops when initializing without lightnvm

Commit Message

Matias Bjørling Feb. 11, 2016, 1:49 p.m. UTC

If the LightNVM subsystem is not compiled into the kernel, and the
null_blk device driver requests lightnvm to be initialized. The call to
nvm_register fails and the null_add_dev function cleans up the
initialization. However, at this point the null block device has
already been added to the nullb_list and thus a second cleanup will
occur when the function has returned, that leads to a double call to
blk_cleanup_queue.

Signed-off-by: Matias Bjørling <[e-mail]>
---
 drivers/block/null_blk.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Comments

Jens Axboe Feb. 11, 2016, 3:56 p.m. UTC | #1

On 02/11/2016 06:49 AM, Matias Bjørling wrote:
> If the LightNVM subsystem is not compiled into the kernel, and the
> null_blk device driver requests lightnvm to be initialized. The call to
> nvm_register fails and the null_add_dev function cleans up the
> initialization. However, at this point the null block device has
> already been added to the nullb_list and thus a second cleanup will
> occur when the function has returned, that leads to a double call to
> blk_cleanup_queue.

Deferring the add seems fine, applied for 4.5.

Patch

diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c
index ae05d31..64a7b59 100644
--- a/drivers/block/null_blk.c
+++ b/drivers/block/null_blk.c
@@ -707,9 +707,7 @@  static int null_add_dev(void)
 	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, nullb->q);
 	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, nullb->q);
 
-
 	mutex_lock(&lock);
-	list_add_tail(&nullb->list, &nullb_list);
 	nullb->index = nullb_indexes++;
 	mutex_unlock(&lock);
 
@@ -743,6 +741,10 @@  static int null_add_dev(void)
 	strncpy(disk->disk_name, nullb->disk_name, DISK_NAME_LEN);
 
 	add_disk(disk);
+
+	mutex_lock(&lock);
+	list_add_tail(&nullb->list, &nullb_list);
+	mutex_unlock(&lock);
 done:
 	return 0;