[PATCH RFC] btrfs: reflink: disable cross-subvolume clone/dedupe for simple quota

[Qu Wenruo <wqu@suse.com>]

Unlike the full qgroup, simple quota no longer makes backref walk to
properly make accurate accounting for each subvolume.

Instead it goes a much faster and much simpler way, anything modified by
the subvolume would be accounted to that subvolume.

Although it brings some small accuracy problem, mostly related to shared
extents between different subvolumes, the reduced overhead is more than
good enough.

Considering there are only 2 ways to share extents between subvolumes:

- Snapshotting
- Cross-subvolume clone/dedupe

And since snapshotting is the core functionality of btrfs, we will never
disable that.

But on the other hand, cross-subvolume snapshotting is not so critical,
and disabling that for simple quota would improve the accuracy of it,
I'd say it's worthy to do that.

Signed-off-by: Qu Wenruo <wqu@suse.com>
---
 fs/btrfs/reflink.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
---
[REASON FOR RFC]
I'm not sure how important the cross-subvolume clone functionality is in
real-world.

But considering squota is mostly designed for container usage, in that
case disabling cross-subvolume clone should be completely fine.

diff --git a/fs/btrfs/reflink.c b/fs/btrfs/reflink.c
index 08d0fb46ceec..906cb6166b67 100644
--- a/fs/btrfs/reflink.c
+++ b/fs/btrfs/reflink.c
@@ -15,6 +15,7 @@
 #include "file-item.h"
 #include "file.h"
 #include "super.h"
+#include "qgroup.h"
 
 #define BTRFS_MAX_DEDUPE_LEN	SZ_16M
 
@@ -350,6 +351,19 @@ static int btrfs_clone(struct inode *src, struct inode *inode,
 	u64 last_dest_end = destoff;
 	u64 prev_extent_end = off;
 
+	/*
+	 * If squota is enabled, disable cloning between different subvolumes.
+	 *
+	 * As clone/reflink/dedupe is the only other way to share data between
+	 * different subvolumes other than snapshotting.
+	 * With it disabled, squota can be way more accurate.
+	 */
+	if (btrfs_qgroup_mode(fs_info) == BTRFS_QGROUP_MODE_SIMPLE) {
+		if (BTRFS_I(src)->root->root_key.objectid !=
+		    BTRFS_I(inode)->root->root_key.objectid)
+			return -EXDEV;
+	}
+
 	ret = -ENOMEM;
 	buf = kvmalloc(fs_info->nodesize, GFP_KERNEL);
 	if (!buf)
-- 
2.44.0