[Intel-gfx] [PATCH 09/14] drm/i915: Introduce i9xx_check_plane_surface()

Ville Syrjälä ville.syrjala at linux.intel.com
Mon Mar 20 17:07:55 UTC 2017


On Fri, Mar 17, 2017 at 10:04:32PM +0000, Chris Wilson wrote:
> On Fri, Mar 17, 2017 at 11:18:03PM +0200, ville.syrjala at linux.intel.com wrote:
> > From: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > 
> > Extract the primary plane surfae offset/x/y calculations for
> > pre-SKL platforms into a common function, and call it during the
> > atomic check phase to reduce the amount of stuff we have to do
> > during the commit phase. SKL is already doing this.
> > 
> > Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > ---
> >  drivers/gpu/drm/i915/intel_display.c | 82 ++++++++++++++++++++++--------------
> >  1 file changed, 50 insertions(+), 32 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
> > index 2e0106a11f8f..024614cb47b6 100644
> > --- a/drivers/gpu/drm/i915/intel_display.c
> > +++ b/drivers/gpu/drm/i915/intel_display.c
> > @@ -3026,6 +3026,43 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
> >  	return dspcntr;
> >  }
> >  
> > +static int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
> > +{
> > +	struct drm_i915_private *dev_priv =
> > +		to_i915(plane_state->base.plane->dev);
> > +	int src_x = plane_state->base.src.x1 >> 16;
> > +	int src_y = plane_state->base.src.y1 >> 16;
> > +	u32 offset;
> > +
> > +	intel_add_fb_offsets(&src_x, &src_y, plane_state, 0);
> > +
> > +	if (INTEL_GEN(dev_priv) >= 4)
> > +		offset = intel_compute_tile_offset(&src_x, &src_y,
> > +						   plane_state, 0);
> > +	else
> > +		offset = 0;
> > +
> > +	/* HSW+ does this automagically in hardware */
> > +	if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)) {
> 
> if (INTEL_GEN() <= 7 && !IS_HASWELL()) {
> 
> would match the comment better.

That would leave out CHV.

I think 'HAS_GMCH || IS_GEN5 || IS_GEN6 || IS_IVB' might be
a semi-decent way to put this. But it's still not quite as
succinct as '!HSW && !BDW'.

What about if I just change the comment to "HSW/BDW do this ..."?

> 
> > +		unsigned int rotation = plane_state->base.rotation;
> > +		int src_w = drm_rect_width(&plane_state->base.src) >> 16;
> > +		int src_h = drm_rect_height(&plane_state->base.src) >> 16;
> > +
> > +		if (rotation & DRM_ROTATE_180) {
> > +			src_x += src_w - 1;
> > +			src_y += src_h - 1;
> > +		} else if (rotation & DRM_REFLECT_X) {
> > +			src_x += src_w - 1;
> > +		}
> > +	}
> > +
> > +	plane_state->main.offset = offset;
> > +	plane_state->main.x = src_x;
> > +	plane_state->main.y = src_y;
> 
> plane_state->actual.offset, ->actual.x, ->actual.y ?
> plane_state->commit.offset, ->commit.x, ->commit.y ?
> 
> Movement looks fine,
> Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk>
> -Chris
> 
> -- 
> Chris Wilson, Intel Open Source Technology Centre

-- 
Ville Syrjälä
Intel OTC


More information about the Intel-gfx mailing list