[Intel-gfx] [PATCH] drm/i915: Fix legacy cursor vs. watermarks for ILK-BDW

[Ville Syrjälä]

On Fri, Feb 17, 2017 at 09:04:44PM +0100, Uwe Kleine-König wrote:
> Hello Ville,
> 
> On Fri, Feb 17, 2017 at 05:01:59PM +0200, ville.syrjala at linux.intel.com wrote:
> > From: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > 
> > In order to make cursor updates actually safe wrt. watermark programming
> > we have to clear the legacy_cursor_update flag in the atomic state. That
> > will cause the regular atomic update path to do the necessary vblank
> > wait after the plane update if needed, otherwise the vblank wait would
> > be skipped and we'd feed the optimal watermarks to the hardware before
> > the plane update has actually happened.
> > 
> > [...]
> > 
> > Cc: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
> > Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> > Cc: Uwe Kleine-König <uwe at kleine-koenig.org>
> > Reported-by: Uwe Kleine-König <uwe at kleine-koenig.org>
> > Fixes: f79f26921ee1 ("drm/i915: Add a cursor hack to allow converting legacy page flip to atomic, v3.")
> > Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> 
> Is this supposed to fix
> https://bugs.freedesktop.org/show_bug.cgi?id=98742 ? If so, the Fixes:
> line seems wrong because f79f26921ee1 isn't in 4.9 where I see the
> issue.
> 
> If I want to fix 4.9---my ultimate goal is to fix the kernel that will
> go into the next stable release---I have to cherry-pick f79f26921ee1
> first, I assume?

Argh. The fact is that it has actually been broken since forever,
but I suppose *something* must have been masking the problem for
you on earlier kernels, well, assuming you didn't actually hit the
problem before 4.9.

You can give backporting the custom legacy cursor path a try,
but I'm not sure how badly it depends on other things. So not
sure a backport is entirely trivial. Maarten, any thoughts?

-- 
Ville Syrjälä
Intel OTC