Re: [PATCH] Btrfs: fix enospc in punch hole

[Filipe Manana <fdmanana@gmail.com>]

On Tue, Oct 11, 2016 at 9:58 AM, robbieko <robbieko@synology.com> wrote:
> Hi Filipe:
>
> because btrfs_calc_trunc_metadata_size is reserved leafsize + nodesize * (8
> - 1)
> assume leafsize is the same as nodesize

The leaf size is always the same as the node size. There are no exceptions.

> , we total reserved 8 nodesize
> when split leaf, we need 2 path

2 paths?? what do you mean?

> , if extent_tree level small than 4, it's OK
> because worst case is (leafsize + nodesize * 3) *2, is 8 nodesize.
> but if extent_tree is greater level 4, worst case is need (leafsize +
> nodesize * 7) * 2,
> is bigger than resvered size, so we should use
> btrfs_calc_trans_metadata_size,
> is taken into account split leaf case.

I think I can make some sense of what you're saying, but you forgot to
mention that after splitting a leaf (therefore creating a new one), a
new node might be added to each level of the tree (since there's a new
key and every parent node was full).

Having detailed and well written change logs is important...

Thanks


>
> Thanks.
> robbieko
>
> Filipe Manana 於 2016-10-07 18:18 寫到:
>
>> On Fri, Oct 7, 2016 at 7:09 AM, robbieko <robbieko@synology.com> wrote:
>>>
>>> From: Robbie Ko <robbieko@synology.com>
>>>
>>> when extent-tree level > BTRFS_MAX_LEVEL / 2,
>>> __btrfs_drop_extents -> btrfs_duplicate_item ->
>>> setup_leaf_for_split -> split_leaf
>>> maybe enospc, because min_size is too small,
>>> need use btrfs_calc_trans_metadata_size.
>>
>>
>> This change log is terrible.
>> You should describe the problem and fix. That is, that hole punching
>> can result in adding new leafs (and as a consequence new nodes) to the
>> tree because when we find file extent items that span beyond the hole
>> range we may end up not deleting them (just adjusting them) and add
>> new file extent items representing holes.
>>
>> And I don't see why this is exclusive for the case where the height of
>> the extent tree is greater than 4 (BTRFS_MAX_LEVEL / 2).
>>
>> The code changes themselves look good to me.
>>
>> thanks
>>
>>>
>>> Signed-off-by: Robbie Ko <robbieko@synology.com>
>>> ---
>>>  fs/btrfs/file.c | 4 ++--
>>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
>>> index fea31a4..809ca85 100644
>>> --- a/fs/btrfs/file.c
>>> +++ b/fs/btrfs/file.c
>>> @@ -2322,7 +2322,7 @@ static int btrfs_punch_hole(struct inode *inode,
>>> loff_t offset, loff_t len)
>>>         u64 tail_len;
>>>         u64 orig_start = offset;
>>>         u64 cur_offset;
>>> -       u64 min_size = btrfs_calc_trunc_metadata_size(root, 1);
>>> +       u64 min_size = btrfs_calc_trans_metadata_size(root, 1);
>>>         u64 drop_end;
>>>         int ret = 0;
>>>         int err = 0;
>>> @@ -2469,7 +2469,7 @@ static int btrfs_punch_hole(struct inode *inode,
>>> loff_t offset, loff_t len)
>>>                 ret = -ENOMEM;
>>>                 goto out_free;
>>>         }
>>> -       rsv->size = btrfs_calc_trunc_metadata_size(root, 1);
>>> +       rsv->size = btrfs_calc_trans_metadata_size(root, 1);
>>>         rsv->failfast = 1;
>>>
>>>         /*
>>> --
>>> 1.9.1
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>



-- 
Filipe David Manana,

"People will forget what you said,
 people will forget what you did,
 but people will never forget how you made them feel."