blksnap: no feedback received on latest patchset - any concerns?

blksnap: no feedback received on latest patchset - any concerns?

[Fabio Fantoni]

Hi Jens, hi all,

The blksnap v7 patchset was posted in February 2024 and has received no 
response so far. As someone who has followed this project closely, I 
wanted to ask directly: is there a reason this patchset is not being 
considered?

This is a module for non-persistent block device snapshots, useful for 
backup on filesystems like ext4 that don't have their own snapshot 
mechanism. It went through 7 revision cycles with review of some people 
and substantial contributions from Christoph Hellwig. The amount of work 
that Sergei Shtepa and Christoph put into this has been remarkable.

The silence after v7 seems to have led the developer to abandon the 
upstream effort entirely, which is unfortunate because seems to me 
there's still a need for a block-level snapshot capabilities module 
integrated upstream, or am I wrong?

I fully understand if there are technical objections, but could you 
please share what they are? Even a definitive rejection, accompanied by 
clear reasoning, would help the community understand whether it’s better 
to take a different approach or address specific concerns in similar cases.

Thank you for your time.

Re: blksnap: no feedback received on latest patchset - any concerns?

[Christoph Hellwig]

Hi Fabio,

a 2 year old patchset has been paged out of everyones mind.  I still
think functionality like it would be useful, but it needs to be resend
to get fresh eyes.

On Thu, Mar 19, 2026 at 01:06:10PM +0100, Fabio Fantoni wrote:
> Hi Jens, hi all,
> 
> The blksnap v7 patchset was posted in February 2024 and has received no
> response so far. As someone who has followed this project closely, I wanted
> to ask directly: is there a reason this patchset is not being considered?
> 
> This is a module for non-persistent block device snapshots, useful for
> backup on filesystems like ext4 that don't have their own snapshot
> mechanism. It went through 7 revision cycles with review of some people and
> substantial contributions from Christoph Hellwig. The amount of work that
> Sergei Shtepa and Christoph put into this has been remarkable.
> 
> The silence after v7 seems to have led the developer to abandon the upstream
> effort entirely, which is unfortunate because seems to me there's still a
> need for a block-level snapshot capabilities module integrated upstream, or
> am I wrong?
> 
> I fully understand if there are technical objections, but could you please
> share what they are? Even a definitive rejection, accompanied by clear
> reasoning, would help the community understand whether it’s better to take a
> different approach or address specific concerns in similar cases.
> 
> Thank you for your time.
> 
---end quoted text---

Re: blksnap: no feedback received on latest patchset - any concerns?

[Jens Axboe]

On 3/19/26 10:16 AM, Christoph Hellwig wrote:
> Hi Fabio,
> 
> a 2 year old patchset has been paged out of everyones mind.  I still
> think functionality like it would be useful, but it needs to be resend
> to get fresh eyes.

Indeed, it's ancient and nobody remembers any of that anymore. If
someone wants to carry it forward and repost, we can take another look.
I wasn't personally convinced about it, and I'm always worried about big
patchsets like that being dump-and-runs - once you've achieved the
company goal of getting it upstream, motivation to maintain it and work
on it often goes away pretty quickly. And then we're left with the pile
of code.

-- 
Jens Axboe