Re: [PATCH] dell-laptop: Adds support for keyboard backlight timeout AC settings

[Darren Hart]

On Tue, Apr 25, 2017 at 05:50:57PM -0400, Arcadiy Ivanov wrote:
> On 2017-04-25 09:49, Mario.Limonciello@xxxxxxxx wrote:
> > > -----Original Message-----
> > > From: Pali Rohár [mailto:pali.rohar@xxxxxxxxx]
> > > Sent: Tuesday, April 25, 2017 8:41 AM
> > > To: Andy Shevchenko <andy.shevchenko@xxxxxxxxx>
> > > Cc: Matthew Garrett <mjg59@xxxxxxxxxxxxx>; Darren Hart
> > > <dvhart@xxxxxxxxxxxxx>; Andy Shevchenko <andy@xxxxxxxxxxxxx>; Arcadiy Ivanov
> > > <arcadiy@xxxxxxxxxx>; Limonciello, Mario <Mario_Limonciello@xxxxxxxx>;
> > > Platform Driver <platform-driver-x86@xxxxxxxxxxxxxxx>; linux-
> > > kernel@xxxxxxxxxxxxxxx
> > > Subject: Re: [PATCH] dell-laptop: Adds support for keyboard backlight timeout AC
> > > settings
> > > 
> > > On Tuesday 25 April 2017 16:32:36 Andy Shevchenko wrote:
> > > > On Tue, Apr 25, 2017 at 3:54 PM, Pali Rohár <pali.rohar@xxxxxxxxx> wrote:
> > > > > On Tuesday 25 April 2017 15:36:56 Andy Shevchenko wrote:
> > > > > > On Sun, Apr 23, 2017 at 10:40 PM, Pali Rohár <pali.rohar@xxxxxxxxx> wrote:
> > > > > > This patch misses few others that were applied to our testing / for-next.
> > > > > Ah... sorry for that :-( And I was in impression that there were fixes
> > > > > for that code which use mutexes... Now I know why I have not seen them!
> > > > > 
> > > > > > I have applied it to testing with corrected conflict (easy to fix),
> > > > > > though check if everything still as supposed to be.
> > > > > Can you send a link to your fix?
> > > > Sure, here it is:
> > > > http://git.infradead.org/linux-platform-drivers-
> > > x86.git/commit/bcf7d8a30e2888f78a19778a16d3dd8c10b4b0ad
> > > 
> > > It is OK.
> > > 
> > Pali,
> > 
> > Considering this is negatively affecting folks on stable kernel releases too
> > when using this newer HW, would you consider to send to @stable too?
> > 
> > Thanks,
> 
> I second that. Thanks!
> 

In order for this to go back to stable, we will need to include the list of
dependencies per stable-kernel-rules.txt. This can be done after it is merged,
but if you can provide the set of Cc: stable lines needed, we can look at
getting this into testing now so no further action will be required.

...

Upon closer inspection, this does fail the 100 lines including context stable
rule by an additional 55 lines. This is a process issue I've been thinking about
for a while and trying to find the best way to express this problem and what a
viable solution might look like. As it stands, fixes such as these are
effectively limited to current and future kernel versions. I do think there is a
need for vendors to be able to update their drivers upstream to work with current
hardware without requiring a bleeding edge kernel. Perhaps "leaf node drivers"
should have a looser set of stable rules, but that is not currently the way the
upstream stable kernel rules work (at least as far as I understand them).

+Greg KH who has the ultimate say on this. Greg, have I misrepresented anything
here?

-- 
Darren Hart
VMware Open Source Technology Center