[Intel-gfx] [PATCH] drm/i915: Do not update pipe state when crtc is inactive.

[Daniel Vetter]

On Wed, Sep 23, 2015 at 01:33:29PM +0300, Jani Nikula wrote:
> On Wed, 23 Sep 2015, Maarten Lankhorst <maarten.lankhorst at linux.intel.com> wrote:
> > Op 23-09-15 om 11:01 schreef Jani Nikula:
> >> On Tue, 22 Sep 2015, Maarten Lankhorst <maarten.lankhorst at linux.intel.com> wrote:
> >>> Nothing good can come from detaching scalers or updating pipe config
> >>> when the crtc is already disabled. Touching registers while the crtc
> >>> and power wells are disabled causes unclaimed register access warnings.
> >> -fixes maintainer grumble. How am I supposed to decipher from this
> >> Subject: line and commit message whether this is a real fix to a real
> >> problem out there, and therefore to be queued for current -rc
> >> development kernels and possibly cc: stable, or not?
> >>
> >> When in doubt, I usually shrug it off, and decide it's for
> >> drm-intel-next-queued, and therefore SEP.
> >>
> > It's for v4.3, so -fixes. Should I explicitly mention it in the commit
> > message next time?
> 
> If you have that information, please do. If not in the commit message,
> then either as a comment after the --- line or in the subject as [PATCH
> for v4.3] or something. Otherwise you'll be relying on my competence too
> much. ;)

What you really should do is mention which commit broke stuff and if
applicable, add a Bugzilla: link. That's required information anyway for
regression fixes, and we have maintainer scripts to tell us where to place
the patch.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch