Re: [PATCH v2 3/3] btrfs: update stripe_extent delete loop assumptions

[Qu Wenruo]

在 2024/7/11 16:25, Qu Wenruo 写道:
> 在 2024/7/11 15:51, Johannes Thumshirn 写道:
> > From: Johannes Thumshirn <johannes.thumshirn@xxxxxxx>
> >
> > btrfs_delete_raid_extent() was written under the assumption, that it's
> > call-chain always passes a start, length tuple that matches a single
> > extent. But btrfs_delete_raid_extent() is called by
> > do_free_extent_acounting() which in term is called by > 
> > __btrfs_free_extent().
>
> But from the call site __btrfs_free_extent(), it is still called for a 
> single extent.
>
> Or we will get an error and abort the current transaction.

Or does it mean, one data extent can have multiple RST entries?

Is that a non-zoned RST specific behavior?
As I still remember that we split ordered extents for zoned devices, so 
that it should always have one extent for each split OE.

Thanks,
Qu
> > But this call-chain passes in a start address and a length that can
> > possibly match multiple on-disk extents.
>
> Mind to give a more detailed example on this?
>
> Thanks,
> Qu
>
> > To make this possible, we have to adjust the start and length of each
> > btree node lookup, to not delete beyond the requested range.
> >
> > Signed-off-by: Johannes Thumshirn <johannes.thumshirn@xxxxxxx>
> > ---
> >   fs/btrfs/raid-stripe-tree.c | 5 +++++
> >   1 file changed, 5 insertions(+)
> >
> > diff --git a/fs/btrfs/raid-stripe-tree.c b/fs/btrfs/raid-stripe-tree.c
> > index fd56535b2289..6f65be334637 100644
> > --- a/fs/btrfs/raid-stripe-tree.c
> > +++ b/fs/btrfs/raid-stripe-tree.c
> > @@ -66,6 +66,11 @@ int btrfs_delete_raid_extent(struct 
> > btrfs_trans_handle *trans, u64 start, u64 le
> >           if (ret)
> >               break;
> > +        start += key.offset;
> > +        length -= key.offset;
> > +        if (length == 0)
> > +            break;
> > +
> >           btrfs_release_path(path);
> >       }