loop: make autoclear operation asynchronous

syzbot is reporting circular locking problem at __loop_clr_fd() [1], for
commit 87579e9b7d8dc36e ("loop: use worker per cgroup instead of kworker")
is calling destroy_workqueue() with disk->open_mutex held.

This circular dependency cannot be broken unless we call __loop_clr_fd()
without holding disk->open_mutex. Therefore, defer __loop_clr_fd() from
lo_release() to a WQ context.

Link: https://syzkaller.appspot.com/bug?extid=643e4ce4b6ad1347d372
Reported-by: syzbot <syzbot+643e4ce4b6ad1347d372@syzkaller.appspotmail.com>
Suggested-by: Christoph Hellwig <hch@infradead.org>
Cc: Jan Kara <jack@suse.cz>
Tested-by: syzbot+643e4ce4b6ad1347d372@syzkaller.appspotmail.com
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/1ed7df28-ebd6-71fb-70e5-1c2972e05ddb@i-love.sakura.ne.jp
Signed-off-by: Jens Axboe <axboe@kernel.dk>

diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index ba76319b554483c001fc8cb73bb02de67262a379..7f4ea06534c2d5f6069dbff7a4c286e751f59de0 100644 (file)
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -1082,7 +1082,7 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
        return error;
 }
 
-static void __loop_clr_fd(struct loop_device *lo, bool release)
+static void __loop_clr_fd(struct loop_device *lo)
 {
        struct file *filp;
        gfp_t gfp = lo->old_gfp_mask;
@@ -1144,8 +1144,6 @@ static void __loop_clr_fd(struct loop_device *lo, bool release)
        /* let user-space know about this change */
        kobject_uevent(&disk_to_dev(lo->lo_disk)->kobj, KOBJ_CHANGE);
        mapping_set_gfp_mask(filp->f_mapping, gfp);
-       /* This is safe: open() is still holding a reference. */
-       module_put(THIS_MODULE);
        blk_mq_unfreeze_queue(lo->lo_queue);
 
        disk_force_media_change(lo->lo_disk, DISK_EVENT_MEDIA_CHANGE);
@@ -1153,44 +1151,52 @@ static void __loop_clr_fd(struct loop_device *lo, bool release)
        if (lo->lo_flags & LO_FLAGS_PARTSCAN) {
                int err;
 
-               /*
-                * open_mutex has been held already in release path, so don't
-                * acquire it if this function is called in such case.
-                *
-                * If the reread partition isn't from release path, lo_refcnt
-                * must be at least one and it can only become zero when the
-                * current holder is released.
-                */
-               if (!release)
-                       mutex_lock(&lo->lo_disk->open_mutex);
+               mutex_lock(&lo->lo_disk->open_mutex);
                err = bdev_disk_changed(lo->lo_disk, false);
-               if (!release)
-                       mutex_unlock(&lo->lo_disk->open_mutex);
+               mutex_unlock(&lo->lo_disk->open_mutex);
                if (err)
                        pr_warn("%s: partition scan of loop%d failed (rc=%d)\n",
                                __func__, lo->lo_number, err);
                /* Device is gone, no point in returning error */
        }
 
-       /*
-        * lo->lo_state is set to Lo_unbound here after above partscan has
-        * finished. There cannot be anybody else entering __loop_clr_fd() as
-        * Lo_rundown state protects us from all the other places trying to
-        * change the 'lo' device.
-        */
        lo->lo_flags = 0;
        if (!part_shift)
                lo->lo_disk->flags |= GENHD_FL_NO_PART;
+
+       fput(filp);
+}
+
+static void loop_rundown_completed(struct loop_device *lo)
+{
        mutex_lock(&lo->lo_mutex);
        lo->lo_state = Lo_unbound;
        mutex_unlock(&lo->lo_mutex);
+       module_put(THIS_MODULE);
+}
 
-       /*
-        * Need not hold lo_mutex to fput backing file. Calling fput holding
-        * lo_mutex triggers a circular lock dependency possibility warning as
-        * fput can take open_mutex which is usually taken before lo_mutex.
-        */
-       fput(filp);
+static void loop_rundown_workfn(struct work_struct *work)
+{
+       struct loop_device *lo = container_of(work, struct loop_device,
+                                             rundown_work);
+       struct block_device *bdev = lo->lo_device;
+       struct gendisk *disk = lo->lo_disk;
+
+       __loop_clr_fd(lo);
+       kobject_put(&bdev->bd_device.kobj);
+       module_put(disk->fops->owner);
+       loop_rundown_completed(lo);
+}
+
+static void loop_schedule_rundown(struct loop_device *lo)
+{
+       struct block_device *bdev = lo->lo_device;
+       struct gendisk *disk = lo->lo_disk;
+
+       __module_get(disk->fops->owner);
+       kobject_get(&bdev->bd_device.kobj);
+       INIT_WORK(&lo->rundown_work, loop_rundown_workfn);
+       queue_work(system_long_wq, &lo->rundown_work);
 }
 
 static int loop_clr_fd(struct loop_device *lo)
@@ -1222,7 +1228,8 @@ static int loop_clr_fd(struct loop_device *lo)
        lo->lo_state = Lo_rundown;
        mutex_unlock(&lo->lo_mutex);
 
-       __loop_clr_fd(lo, false);
+       __loop_clr_fd(lo);
+       loop_rundown_completed(lo);
        return 0;
 }
 
@@ -1747,7 +1754,7 @@ static void lo_release(struct gendisk *disk, fmode_t mode)
                 * In autoclear mode, stop the loop thread
                 * and remove configuration after last close.
                 */
-               __loop_clr_fd(lo, true);
+               loop_schedule_rundown(lo);
                return;
        } else if (lo->lo_state == Lo_bound) {
                /*

diff --git a/drivers/block/loop.h b/drivers/block/loop.h
index 082d4b6bfc6a6cbe789fa2c6361ae3a88a8b3e27..918a7a2dc025933a50dfe1ef8bc420b51639a9c0 100644 (file)
--- a/drivers/block/loop.h
+++ b/drivers/block/loop.h
@@ -56,6 +56,7 @@ struct loop_device {
        struct gendisk          *lo_disk;
        struct mutex            lo_mutex;
        bool                    idr_visible;
+       struct work_struct      rundown_work;
 };
 
 struct loop_cmd {