Re: [PATCHv3 0/8] introduce dynamic device creation/removal

[Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>]

Hello,

On (04/16/15 08:40), Minchan Kim wrote:
> On Wed, Apr 15, 2015 at 02:37:17PM -0700, Andrew Morton wrote:
> > On Tue,  3 Mar 2015 21:49:42 +0900 Sergey Senozhatsky <sergey.senozhatsky@gmail.com> wrote:
> > 
> > > Hello,
> > > 
> > > This patchset introduces zram-control sysfs class, which has two sysfs
> > > attrs:
> > >  - zram_add     -- add a new specific (device_id) zram device
> > >  - zram_remove  -- remove a specific (device_id) zram device
> > 
> > This patchset and the "make automatic device_id generation possible"
> > still appear to have quite a few unresolved issues.  So I'm holding
> > them out of the 4.1 merge window.
> 
> There is no unresolved issue to me. Only one thing I suspect was the
> feature user enforce new device id for dynamic device addition and
> we finally decided to remove the function because there was no useful
> usecase at this point.

I'm not aware of any unresolved issues. am I missing something?


> Sergey and other userland people agreed that
> so Sergey sent a patch [zram: do not let user enforce new device dev_id]
> https://lkml.org/lkml/2015/3/6/427
> So, I'm happy with that. Acutally, I wanted to resend whole patchset
> for dynamic device creation/remove patchset with corrected version
> (ie, remove user enforce new device id) to avoid confusion but didn't
> said it to Sergey. It was my bad.
> 
> Sergey, Could you resend this patchset without user's enforce device id
> function based on new -rc1?

ok, agree. I'll re-submit later today.


	-ss

> > 
> > Unfortunately these were the first-arriving zram patches, so the later
> > ones required quite a bit of mangling.  Hopefully I got it all right.
> > 
> > This was all a bit disruptive.  Please let's not leave major patchsets
> > floating about in an incomplete/unresolved state for week after week?
> 
> I will keep it in mind.
> Thanks.
> 
> -- 
> Kind regards,
> Minchan Kim
>