Skip to contentSkip to conversation

fs/block_dev.c: Remove WARN_ON() when inode writeback fails

PATCH In review Jeff Moyer 2 participants 3 messages Oct 30, 2015 – Nov 2, 2015 VFS

lore
Share & tools
lore

Jeff Moyer

author
#0 lore raw

PATCH v2 1/1 fs/block_dev.c: Remove WARN_ON() when inode writeback fails

From: Vivek Goyal <[email protected]>

If a block device is hot removed and later last reference to device is put, we try to writeback the dirty inode. But device is gone and that writeback fails.

Currently we do a WARN_ON() which does not seem to be the right thing. Convert it to a ratelimited kernel warning.

[[email protected]: get rid of unnecessary name initialization, 80 cols]

---

Jens, we're still getting reports of this problem in Fedora:
https://bugzilla.redhat.com/show_bug.cgi?id=1196089

Here's the last posting and discussion of this patch for reference:
https://lkml.org/lkml/2015/6/22/487

fs/block_dev.c +12 -3
index 073bb57..d9d3039 100644
@@ -50,12 +50,21 @@ struct block_device *I_BDEV(struct inode *inode)
 }
 EXPORT_SYMBOL(I_BDEV);
 
-static void bdev_write_inode(struct inode *inode)
+static void bdev_write_inode(struct block_device *bdev)
 {
+	struct inode *inode = bdev->bd_inode;
+	int ret;
+
 	spin_lock(&inode->i_lock);
 	while (inode->i_state & I_DIRTY) {
 		spin_unlock(&inode->i_lock);
-		WARN_ON_ONCE(write_inode_now(inode, true));
+		ret = write_inode_now(inode, true);
+		if (ret) {
+			char name[BDEVNAME_SIZE];
+			pr_warn_ratelimited("VFS: Dirty inode writeback failed "
+					    "for block device %s (err=%d).\n",
+					    bdevname(bdev, name), ret);
+		}
 		spin_lock(&inode->i_lock);
 	}
 	spin_unlock(&inode->i_lock);
@@ -1504,7 +1513,7 @@ static void __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part)
 		 * ->release can cause the queue to disappear, so flush all
 		 * dirty data before.
 		 */
-		bdev_write_inode(bdev->bd_inode);
+		bdev_write_inode(bdev);
 	}
 	if (bdev->bd_contains == bdev) {
 		if (disk->fops->release)
#1 lore raw
Quoted 8 lines · Jeff Moyer

From: Vivek Goyal <[email protected]>

If a block device is hot removed and later last reference to device is put, we try to writeback the dirty inode. But device is gone and that writeback fails.

Currently we do a WARN_ON() which does not seem to be the right thing. Convert it to a ratelimited kernel warning.

Any concerns with putting this into 4.4 and marking it for stable? Would be awesome to have a "Fixes:" in here too.

Signature · 1 line
Jens Axboe
author
#2 lore raw
Quoted 11 lines · Jens Axboe
Quoted 8 lines · Jeff Moyer

From: Vivek Goyal <[email protected]>

If a block device is hot removed and later last reference to device is put, we try to writeback the dirty inode. But device is gone and that writeback fails.

Currently we do a WARN_ON() which does not seem to be the right thing. Convert it to a ratelimited kernel warning.

Any concerns with putting this into 4.4 and marking it for stable?

No, that sounds fine to me.

Quoted 1 line

Would be awesome to have a "Fixes:" in here too.

OK. I don't know which commit introduced the problem. Do you want something like this?

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1196089
or
Fixes: https://www.redhat.com/archives/dm-devel/2015-June/msg00111.html

I know some folks don't like mailing list archive links in the commit messages.

Cheers,
Jeff