[PATCH] Btrfs: fix a bug of snapshot-aware defrag to make it work on partial extents

[PATCH] Btrfs: fix a bug of snapshot-aware defrag to make it work on partial extents

[Liu Bo]

For partial extents, snapshot-aware defrag does not work as expected,
since
a) we use the wrong logical offset to search for parents, which should be
   disk_bytenr + extent_offset, not just disk_bytenr,
b) 'offset' returned by the backref walking just refers to key.offset, not
   the 'offset' stored in btrfs_extent_data_ref which is
   (key.offset - extent_offset).

The reproducer:
$ mkfs.btrfs sda
$ mount sda /mnt
$ btrfs sub create /mnt/sub
$ for i in `seq 5 -1 1`; do dd if=/dev/zero of=/mnt/sub/foo bs=5k count=1 seek=$i conv=notrunc oflag=sync; done
$ btrfs sub snap /mnt/sub /mnt/snap1
$ btrfs sub snap /mnt/sub /mnt/snap2
$ sync; btrfs filesystem defrag /mnt/sub/foo;
$ umount /mnt
$ btrfs-debug-tree sda (Here we can check whether the defrag operation is snapshot-awared.

This addresses the above two problems.

Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
---
 fs/btrfs/inode.c |   16 ++++++++++++----
 1 files changed, 12 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index c931a4d..ab87862 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -2134,16 +2134,23 @@ static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
 		if (btrfs_file_extent_disk_bytenr(leaf, extent) != old->bytenr)
 			continue;
 
-		extent_offset = btrfs_file_extent_offset(leaf, extent);
-		if (key.offset - extent_offset != offset)
+		/*
+		 * 'offset' refers to the exact key.offset,
+		 * NOT the 'offset' field in btrfs_extent_data_ref, ie.
+		 * (key.offset - extent_offset).
+		 */
+		if (key.offset != offset)
 			continue;
 
+		extent_offset = btrfs_file_extent_offset(leaf, extent);
 		num_bytes = btrfs_file_extent_num_bytes(leaf, extent);
+
 		if (extent_offset >= old->extent_offset + old->offset +
 		    old->len || extent_offset + num_bytes <=
 		    old->extent_offset + old->offset)
 			continue;
 
+		ret = 0;
 		break;
 	}
 
@@ -2155,7 +2162,7 @@ static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id,
 
 	backref->root_id = root_id;
 	backref->inum = inum;
-	backref->file_pos = offset + extent_offset;
+	backref->file_pos = offset;
 	backref->num_bytes = num_bytes;
 	backref->extent_offset = extent_offset;
 	backref->generation = btrfs_file_extent_generation(leaf, extent);
@@ -2178,7 +2185,8 @@ static noinline bool record_extent_backrefs(struct btrfs_path *path,
 	new->path = path;
 
 	list_for_each_entry_safe(old, tmp, &new->head, list) {
-		ret = iterate_inodes_from_logical(old->bytenr, fs_info,
+		ret = iterate_inodes_from_logical(old->bytenr +
+						  old->extent_offset, fs_info,
 						  path, record_one_backref,
 						  old);
 		BUG_ON(ret < 0 && ret != -ENOENT);
-- 
1.7.7

Re: [PATCH] Btrfs: fix a bug of snapshot-aware defrag to make it work on partial extents

[Josef Bacik]

On Mon, Jul 01, 2013 at 10:13:26PM +0800, Liu Bo wrote:
> For partial extents, snapshot-aware defrag does not work as expected,
> since
> a) we use the wrong logical offset to search for parents, which should be
>    disk_bytenr + extent_offset, not just disk_bytenr,
> b) 'offset' returned by the backref walking just refers to key.offset, not
>    the 'offset' stored in btrfs_extent_data_ref which is
>    (key.offset - extent_offset).
> 
> The reproducer:
> $ mkfs.btrfs sda
> $ mount sda /mnt
> $ btrfs sub create /mnt/sub
> $ for i in `seq 5 -1 1`; do dd if=/dev/zero of=/mnt/sub/foo bs=5k count=1 seek=$i conv=notrunc oflag=sync; done
> $ btrfs sub snap /mnt/sub /mnt/snap1
> $ btrfs sub snap /mnt/sub /mnt/snap2
> $ sync; btrfs filesystem defrag /mnt/sub/foo;
> $ umount /mnt
> $ btrfs-debug-tree sda (Here we can check whether the defrag operation is snapshot-awared.
> 
> This addresses the above two problems.
> 

Can this be turned into a xfstest somehow?  Like do fiemap and make sure the
block numbers match up right?  Thanks,

Josef

Re: [PATCH] Btrfs: fix a bug of snapshot-aware defrag to make it work on partial extents

[Liu Bo]

On Mon, Jul 01, 2013 at 04:22:06PM -0400, Josef Bacik wrote:
> On Mon, Jul 01, 2013 at 10:13:26PM +0800, Liu Bo wrote:
> > For partial extents, snapshot-aware defrag does not work as expected,
> > since
> > a) we use the wrong logical offset to search for parents, which should be
> >    disk_bytenr + extent_offset, not just disk_bytenr,
> > b) 'offset' returned by the backref walking just refers to key.offset, not
> >    the 'offset' stored in btrfs_extent_data_ref which is
> >    (key.offset - extent_offset).
> > 
> > The reproducer:
> > $ mkfs.btrfs sda
> > $ mount sda /mnt
> > $ btrfs sub create /mnt/sub
> > $ for i in `seq 5 -1 1`; do dd if=/dev/zero of=/mnt/sub/foo bs=5k count=1 seek=$i conv=notrunc oflag=sync; done
> > $ btrfs sub snap /mnt/sub /mnt/snap1
> > $ btrfs sub snap /mnt/sub /mnt/snap2
> > $ sync; btrfs filesystem defrag /mnt/sub/foo;
> > $ umount /mnt
> > $ btrfs-debug-tree sda (Here we can check whether the defrag operation is snapshot-awared.
> > 
> > This addresses the above two problems.
> > 
> 
> Can this be turned into a xfstest somehow?  Like do fiemap and make sure the
> block numbers match up right?  Thanks,

Yeah, I've sent a patch for this :)

- liubo