Re: [PATCH 0/2] btrfs: defrag: don't defrag extents already at their max capacity, then remove an ambiguous check

[Filipe Manana <fdmanana@kernel.org>]

On Thu, Jan 27, 2022 at 01:24:41PM +0800, Qu Wenruo wrote:
> Thanks to the report from Filipe where he found a new bug in compressed
> defrag, that we will try to defrag all compressed extents even it's
> already at its max compacity.
> 
> This behavior is from the beginning of btrfs defrag.
> 
> The first patch is to fix the behavior, by just rejecting extents
> which are already at their max capacity, nor allowing extents to be
> merged with extents at their max capacity.
> 
> The second patch is to remove an ambiguous rejection condition.
> The condition is believed to reject compressed extent, but it never
> really works due to the check is > 128K, not >= 128K.
> 
> And the physically adajcent check may prevent users to reduce the number
> of extents.

With so many patches sent in such short periods of time, with some in series
and others not in properly formatted series, it's likely hard for other people
to follow what's going on.

So this patchset, applies on top of the following patch:

https://patchwork.kernel.org/project/linux-btrfs/patch/20220126005850.14729-1-wqu@suse.com/

That patch is part of another series consisting of 3 patches, but this
patchset here makes the patches 2/3 and 3/3 of that other patchset
obsolete.

Thanks.

> 
> Qu Wenruo (2):
>   btrfs: defrag: don't defrag extents which is already at its max
>     capacity
>   btrfs: defrag: remove an ambiguous condition for rejection
> 
>  fs/btrfs/ioctl.c | 22 +++++++++++++++++++---
>  1 file changed, 19 insertions(+), 3 deletions(-)
> 
> -- 
> 2.34.1
>