Re: [PATCH 07/18] btrfs: create qgroup earlier in snapshot creation

[Boris Burkov <boris@bur.io>]

On Thu, Jul 13, 2023 at 04:37:04PM -0400, Josef Bacik wrote:
> On Thu, Jul 13, 2023 at 12:00:42PM -0700, Boris Burkov wrote:
> > On Thu, Jul 13, 2023 at 10:26:00AM -0400, Josef Bacik wrote:
> > > On Wed, Jul 05, 2023 at 04:20:44PM -0700, Boris Burkov wrote:
> > > > Pull creating the qgroup earlier in the snapshot. This allows simple
> > > > quotas qgroups to see all the metadata writes related to the snapshot
> > > > being created and to be born with the root node accounted.
> > > > 
> > > > Signed-off-by: Boris Burkov <boris@bur.io>
> > > > ---
> > > >  fs/btrfs/qgroup.c      | 3 +++
> > > >  fs/btrfs/transaction.c | 5 +++++
> > > >  2 files changed, 8 insertions(+)
> > > > 
> > > > diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
> > > > index 75afd8212bc0..8419270f7417 100644
> > > > --- a/fs/btrfs/qgroup.c
> > > > +++ b/fs/btrfs/qgroup.c
> > > > @@ -1670,6 +1670,9 @@ int btrfs_create_qgroup(struct btrfs_trans_handle *trans, u64 qgroupid)
> > > >  	struct btrfs_qgroup *qgroup;
> > > >  	int ret = 0;
> > > >  
> > > > +	if (btrfs_qgroup_mode(fs_info) == BTRFS_QGROUP_MODE_DISABLED)
> > > > +		return 0;
> > > > +
> > > >  	mutex_lock(&fs_info->qgroup_ioctl_lock);
> > > >  	if (!fs_info->quota_root) {
> > > >  		ret = -ENOTCONN;
> > > > diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
> > > > index f644c7c04d53..2bb5a64f6d84 100644
> > > > --- a/fs/btrfs/transaction.c
> > > > +++ b/fs/btrfs/transaction.c
> > > > @@ -1716,6 +1716,11 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
> > > >  	}
> > > >  	btrfs_release_path(path);
> > > >  
> > > > +	ret = btrfs_create_qgroup(trans, objectid);
> > > > +	if (ret) {
> > > > +		btrfs_abort_transaction(trans, ret);
> > > > +		goto fail;
> > > > +	}
> > > 
> > > Newline please.
> > > 
> > > How is this ok with normal qgroups?  We weren't creating a qgroup at snapshot
> > > creation time at all it seems, so I don't understand how this is ok for qgroups.
> > 
> > qgroup_account_snapshot calls btrfs_qgroup_inherit which contains a
> > separate implementation of qgroup creation.
> 
> Which it still does, so I'm confused as to how this is ok.  How do we not get an
> EEXIST when we do the btrfs_qgroup_inherit?  Thanks,
> 
> Josef

add_qgroup_item ignores EEXIST. I believe this should be exercised by
a user intentionally creating a qgroup for a subvolume before
creating the subvolume (I believe that is explicitly supported, though
it does seem to rely on guessing the subvol ID?!)

from man btrfs-qgroup:
"For the 0/<subvolume id> qgroup, a qgroup can be created even before
the subvolume is created."