Unneeded commits in 3.16-ckt branch?

Unneeded commits in 3.16-ckt branch?

[Ben Hutchings]

[-- Attachment #1: Type: text/plain, Size: 1411 bytes --]

So far as I can see, the following commits don't fix bugs in 3.16.  But
I don't know that they cause any problems, either.

commit aba636b877ec270075cbb01b4fceb460f0c3f07a
Author: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
Date:   Thu Apr 16 13:05:19 2015 +0100

    i2c: Mark adapter devices with pm_runtime_no_callbacks
    
    commit 6ada5c1e1b077ab98fc144d7ac132b4dcc0148ec upstream.

- fixes regression in 4.0

commit 65342eab7baa43fb7852aafbe99e54daf2cd6420
Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Date:   Tue Apr 14 15:42:30 2015 -0700

    mm/hugetlb: use pmd_page() in follow_huge_pmd()
    
    commit 97534127012f0e396eddea4691f4c9b170aed74b upstream.

- fixes regression in 4.0

commit 9c518c0eac18b0f0e13666ae67aa0a7e87a81032
Author: Eric Dumazet <edumazet@google.com>
Date:   Thu Apr 9 13:31:56 2015 -0700

    tcp: tcp_make_synack() should clear skb->tstamp
    
    commit b50edd7812852d989f2ef09dcfc729690f54a42d upstream.

- fixes regression in 3.18

commit 46f080a6f59fd58de1acb4ccadf3a8b5f934659a
Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date:   Tue Apr 21 09:49:11 2015 -0700

    drm/i915: cope with large i2c transfers
    
    commit 9535c4757b881e06fae72a857485ad57c422b8d2 upstream.

- needed for feature introduced in 3.17

-- 
Ben Hutchings
Never put off till tomorrow what you can avoid all together.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 811 bytes --]

Re: Unneeded commits in 3.16-ckt branch?

[Luis Henriques]

On Mon, May 18, 2015 at 12:20:23AM +0100, Ben Hutchings wrote:
> So far as I can see, the following commits don't fix bugs in 3.16.  But
> I don't know that they cause any problems, either.
>

Thanks a lot for taking some time reporting this, Ben.  Comments
bellow.

> commit aba636b877ec270075cbb01b4fceb460f0c3f07a
> Author: Charles Keepax <ckeepax@opensource.wolfsonmicro.com>
> Date:   Thu Apr 16 13:05:19 2015 +0100
> 
>     i2c: Mark adapter devices with pm_runtime_no_callbacks
>     
>     commit 6ada5c1e1b077ab98fc144d7ac132b4dcc0148ec upstream.
> 
> - fixes regression in 4.0
> 
> commit 65342eab7baa43fb7852aafbe99e54daf2cd6420
> Author: Gerald Schaefer <gerald.schaefer@de.ibm.com>
> Date:   Tue Apr 14 15:42:30 2015 -0700
> 
>     mm/hugetlb: use pmd_page() in follow_huge_pmd()
>     
>     commit 97534127012f0e396eddea4691f4c9b170aed74b upstream.
> 
> - fixes regression in 4.0
>

For these 2 I will queue a revert for the next 3.16 release.  The
original authors will have a chance to comment if they think they
should be kept.

> commit 9c518c0eac18b0f0e13666ae67aa0a7e87a81032
> Author: Eric Dumazet <edumazet@google.com>
> Date:   Thu Apr 9 13:31:56 2015 -0700
> 
>     tcp: tcp_make_synack() should clear skb->tstamp
>     
>     commit b50edd7812852d989f2ef09dcfc729690f54a42d upstream.
> 
> - fixes regression in 3.18
>

Since this one seems to be harmless (and most of the other stable
kernels have this commit), I don't think it's worth the trouble
reverting it.  In fact, Eric and David commented this patch on 3.10
(or 3.14?) and they agreed it was OK to keep it.

> commit 46f080a6f59fd58de1acb4ccadf3a8b5f934659a
> Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> Date:   Tue Apr 21 09:49:11 2015 -0700
> 
>     drm/i915: cope with large i2c transfers
>     
>     commit 9535c4757b881e06fae72a857485ad57c422b8d2 upstream.
> 
> - needed for feature introduced in 3.17

I agree this one shouldn't have been added to the 3.16 kernel (or any
other kernel < 3.17).  However, it does seem to be harmless and it has
already been included in most of the stable kernels.

3.12 is currently under review and is including this patch too.  I'll
reply to that thread (Cc'ing you), and see if the authors can comment
on this.

Cheers,
--
Lu�s