zram: don't fail to remove zram during unloading module

When the zram module is being unloaded, no one should be using the
zram disks. However even while being unloaded the zram module's
sysfs attributes might be poked at to re-configure zram devices.
This is expected, and kernfs ensures that these operations complete
before device_del() completes.

But reset_store() may set ->claim which will fail zram_remove(), when
this happens, zram_reset_device() is bypassed, and zram->comp can't
be destroyed, so the warning of 'Error: Removing state 63 which has
instances left.' is triggered during unloading module, together with
memory leak and sort of thing.

Fixes the issue by not failing zram_remove() if ->claim is set, and
we actually need to do nothing in case that zram_reset() is running
since del_gendisk() will wait until zram_reset() is done.

Reported-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Luis Chamberlain <mcgrof@kernel.org>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Link: https://lore.kernel.org/r/20211025025426.2815424-3-ming.lei@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 21 insertions, 6 deletions

diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
index 9b38e5f749c6..13b65ebbab8d 100644
--- a/drivers/block/zram/zram_drv.c
+++ b/drivers/block/zram/zram_drv.c
@@ -1972,25 +1972,40 @@ out_free_dev:
 static int zram_remove(struct zram *zram)
 {
 	struct block_device *bdev = zram->disk->part0;
+	bool claimed;
 
 	mutex_lock(&bdev->bd_disk->open_mutex);
-	if (bdev->bd_openers || zram->claim) {
+	if (bdev->bd_openers) {
 		mutex_unlock(&bdev->bd_disk->open_mutex);
 		return -EBUSY;
 	}
 
-	zram->claim = true;
+	claimed = zram->claim;
+	if (!claimed)
+		zram->claim = true;
 	mutex_unlock(&bdev->bd_disk->open_mutex);
 
 	zram_debugfs_unregister(zram);
 
-	/* Make sure all the pending I/O are finished */
-	fsync_bdev(bdev);
-	zram_reset_device(zram);
+	if (claimed) {
+		/*
+		 * If we were claimed by reset_store(), del_gendisk() will
+		 * wait until reset_store() is done, so nothing need to do.
+		 */
+		;
+	} else {
+		/* Make sure all the pending I/O are finished */
+		fsync_bdev(bdev);
+		zram_reset_device(zram);
+	}
 
 	pr_info("Removed device: %s\n", zram->disk->disk_name);
 
 	del_gendisk(zram->disk);
+
+	/* del_gendisk drains pending reset_store */
+	WARN_ON_ONCE(claimed && zram->claim);
+
 	blk_cleanup_disk(zram->disk);
 	kfree(zram);
 	return 0;
@@ -2067,7 +2082,7 @@ static struct class zram_control_class = {
 
 static int zram_remove_cb(int id, void *ptr, void *data)
 {
-	zram_remove(ptr);
+	WARN_ON_ONCE(zram_remove(ptr));
 	return 0;
 }