On Tue, Jul 03, 2012 at 06:04:38AM -0600, Liu Bo wrote:

On 07/03/2012 07:27 PM, Alexander Block wrote:

quoted

On Fri, Jun 29, 2012 at 11:59 AM, Liu Bo [off-list ref] wrote:

quoted

I've modified 'btrfs subvolume list' to show a subvolume's attributes,
such as readonly and default, and adopted a new structure for args for
subvol_getflags/setflags.

So here is the kernel side update.

Signed-off-by: Liu Bo <redacted>
---
 fs/btrfs/ioctl.c |  100 ++++++++++++++++++++++++++++++++++++++++-------------
 fs/btrfs/ioctl.h |    5 +++
 2 files changed, 80 insertions(+), 25 deletions(-)

[...]

quoted

quoted

diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h
index 3186d2d..1fa0ce2 100644
--- a/fs/btrfs/ioctl.h
+++ b/fs/btrfs/ioctl.h

@@ -45,6 +45,11 @@ struct btrfs_ioctl_vol_args_v2 {
        char name[BTRFS_SUBVOL_NAME_MAX + 1];
 };

+struct btrfs_ioctl_get_set_flags_args {
+       __u64 objectid;
+       __u64 flags;
+};
+

Shouldn't BTRFS_IOC_SUBVOL_GETFLAGS/BTRFS_IOC_SUBVOL_SETFLAGS also be
updated? Both still have __u64 as arguments. Also, my patches for the
new btrfs prop command group won't work anymore due to the change in
the ioctl. I'm in the middle of preparing btrfs send/receive to be
sent to the list...I also use the ioctls there and we should agree on
what to do with the ioctls before I send the patches out.

Just drop this patch, I've discussed with Ilya about it (and also CCed you).

If we've more properties pending to set/get, 'btrfs property' will be better.

So please go on your work, I'll try to rebase my patch on yours.

I've dropped the entire series from btrfs-next because without this patch the
other one fails to build without __btrfs_subvol_get_root.  Resend if you want
the other patches included.  Thanks,

Josef